Accelerator for genotype reckoning model

Abstract

This disclosure describes methods, non-transitory computer-readable media, and systems that can determine allele likelihood of genomic regions exhibiting certain haplotype alleles using one or both of merge computing and data exchange across dedicated hardware. For example, the disclosed system may determine an intermediate allele likelihood of a genomic region including haplotype alleles by running a single round of parallel multiplication. In some cases, the disclosed system determines and stores a subset of intermediate allele likelihoods corresponding to a set of tagged variants, and temporarily generates a set of intermediate allele likelihoods of the set of tagged variants by using the subset of intermediate allele likelihoods as a warm start point. In further embodiments, the disclosed system determines a run sum of intermediate allele possibilities for genomic regions exhibiting haplotype alleles for haplotypes given a tagged variant, and using the run sum as an input to determine an intermediate allele likelihood of the genomic region exhibiting the haplotype allele given another tagged variant.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims the benefit of and priority to U.S. Provisional Application No. 63 / 367,105, filed on June 27, 2022, entitled “ACCELERATORS FOR A GENOTYPEIMPUTATION MODEL”, which is hereby incorporated by reference in its entirety. Background Art

[0003] In recent years, biotechnology companies and research institutions have improved the hardware and software platforms for nucleotide sequencing and determining the nucleobase calls of genomic samples. For example, some existing sequencers and sequencing data analysis software (collectively referred to as "existing sequencing systems") determine the individual nucleobases in a sequence by using conventional Sanger sequencing or by using a sequencing-by-synthesis (SBS) method. When using SBS, existing sequencing systems can monitor thousands of oligonucleotides synthesized in parallel from a template to predict the nucleobase calls of growing nucleotide reads. For example, cameras in many existing sequencing systems capture images of irradiated fluorescent labels incorporated into oligonucleotides. After capturing such images, some existing sequencing systems process the image data from the camera and determine the nucleobase calls corresponding to the nucleotide reads of the oligonucleotides. Based on the comparison of the nucleobase calls of such reads with the reference genome, existing systems utilize variant detectors to identify variants in genomic samples, such as single nucleotide polymorphisms (SNPs), insertions or deletions (indels), or other variants in genomic samples.

[0004] Despite these recent advances, existing sequencing systems sometimes fail to accurately determine base calls or collect insufficient (or seemingly contradictory) nucleotide reads, particularly for nucleobases in low read coverage genomic regions. For certain genomic regions of genomic samples, existing sequencing systems typically use genotype inference models to infer nucleobase calls and phase haplotypes based on variants detected in genomic samples. For example, existing sequencing systems often use various types of hidden Markov models (HMMs) customized for inferring genotypes, such as by using genotype likelihood inference and phasing methods (GLIMPSE) or IMPUTE to infer nucleobase calls for specific genomic regions. Although such HMMs have improved the accuracy of inferring genotypes, existing sequencing systems using genotype inference models often consume a lot of computer processing, require a lot of memory to store data generated by the genotype inference model, and execute the genotype inference model with inefficient processor downtime waiting time.

[0005] As just suggested, existing sequencing systems consume too much computer processing and time when performing HMM for genotype inference. For example, some existing sequencing systems running a single thread on a central processing unit (CPU) consume an average of about 17.5 hours to phase and infer the haplotype allele possibilities corresponding to a single marker allele in a genomic region. About 80% of such phasing and inference calculations come from HMM calculations and Burrows-Wheeler transform (BWT), where HMM calculations consume about 60% of the calculation time and BWT calculations consume about 20% of the calculation time. Although BWT calculations may be amortized and significantly reduced in percentage across large batches of genomic samples, HMM calculation time on a single CPU thread may still consume about 10 hours or more (e.g., 600 minutes to 640 minutes).

[0006] In addition to consuming a lot of time and computer processing, existing sequencing systems also consume a lot of memory when executing HMM for genotype inference. For example, in some cases, existing sequencing systems determine and save the value of haplotype allele probability in a haplotype matrix, which corresponds to 50 million cells of a set of marker variants and haplotypes from a haplotype reference group. Given a single haplotype matrix with 50 million cells, an existing sequencing system determines 40,000 haplotype calls based on 40,000 haplotype matrices within a certain period of time, and must determine the value corresponding to 2 trillion cells. Since some HMMs (such as GLIMPSE) used for genotype inference require existing sequencing systems to determine the value once for each haplotype matrix for the α round of the HMM, and twice for the β round, the existing sequencing system can determine and save values ​​across many haplotype matrices of a total of about 6 trillion cells to calculate the HMM-based genotype inference for multiple genomic regions. While hardware on sequencing devices and servers has increased memory, chips used in field-programmable gate arrays (FPGAs) or other configurable processors typically include about 32 or 64 gigabytes of memory on the chip, which is barely enough or insufficient to store the data for a single haplotype matrix.

[0007] In addition to increasing the burden of processing time and memory, some existing sequencing systems are also unable to efficiently perform HMM to perform genotype inference and provide a waiting time cycle for the processor. For example, in some cases, some existing sequencing systems even determine the sum of all intermediate allele possibilities of a marker variant and various haplotypes from a haplotype reference group based on both α round values ​​and β round values ​​before determining the intermediate allele possibilities of subsequent marker variants. Since all intermediate allele possibilities must be summed and the allele possibility of a marker variant must be determined before determining the intermediate allele possibilities of another marker variant, the processor used by the existing sequencing system often waits through a certain waiting time cycle to perform one or both of the following: summing the intermediate allele possibilities of adjacent markers and generating allele possibilities. For genotype inference based on HMM that may require 50 million units of haplotype matrices and 40,000 separate haplotype matrices, this type of calculation waiting time is inefficient and wastes the time that the processor could have been used to further calculate the allele possibility.

[0008] As suggested by the memory size and computation time described above, existing sequencing systems will require significant throughput of input and output values ​​to effectively perform HMM-based genotype inference. Because such inference may require the storage or transmission of large amounts of data (such as certain input values ​​of a haplotype matrix having millions, billions, or tens of thousands of cells), existing HMM-based genotype inference further burdens the bandwidth of high-speed buses (such as the Peripheral Component Interconnect Express (PCIe) or other interfaces that connect processor cards to other hardware within a computing device). Any bottleneck in PCIe throughput or other interface throughput may significantly slow down HMM-based genotype inference.

[0009] These problems and challenges, along with additional problems and challenges, exist with existing sequencing systems. Summary of the invention

[0010] The present disclosure describes one or more embodiments of systems, methods and non-transient computer-readable storage media that solve one or more of the above problems or provide other advantages over the prior art. In order to speed up computer processing or efficiently redistribute the memory load of the genotype inference model, the disclosed system can use merged calculation, efficient data transmission or customized architecture to determine the allele possibility of the genomic region that exhibits certain haplotype alleles. For example, the disclosed system can determine the intermediate allele possibility of the genomic region including the haplotype allele in the case of a given marker variant and a reference group haplotype by a single round of parallel multiplication operations on a processor. In some cases, the disclosed system determines and stores a subset of the intermediate allele possibility corresponding to the marker variant group, and generates a complete intermediate allele possibility set for the marker variant set by using the intermediate allele possibility subset as a hot start point. In a further embodiment, the disclosed system determines a running sum of intermediate allele probabilities for genomic regions that exhibit haplotype alleles for one or more haplotypes given a marker variant, and uses the running sum as a running input to determine individual intermediate allele probabilities for genomic regions that exhibit haplotype alleles given another marker variant, without a waiting time period for summing the intermediate allele probabilities and / or generating the allele probabilities.

[0011] Additional features and advantages of one or more embodiments of the present disclosure will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of such exemplary embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The detailed description refers to the accompanying drawings which are briefly described below.

[0013] Figure 1 An environment in which an accelerated genotype imputation system according to one or more embodiments of the present disclosure may operate is illustrated.

[0014] FIG. 2A to FIG. 2B The invention illustrates that an accelerated genotype imputation system according to one or more embodiments of the present disclosure utilizes a haplotype matrix to execute a hidden Markov model (HMM)-based genotype imputation model to determine the posterior genotype likelihood of genomic regions of a plurality of genomic samples.

[0015] FIG. 3A to FIG. 3B The invention illustrates that an accelerated genotype imputation system according to one or more embodiments of the present disclosure determines an intermediate allele likelihood of a genomic region including haplotype alleles by running a merge operation on a processor.

[0016] FIG. 4A to FIG. 4BIt is illustrated that the accelerated genotype imputation system according to one or more embodiments of the present disclosure determines and stores an intermediate allele likelihood subset as a hot start point, and generates an intermediate allele likelihood set for a marker variant set by using the intermediate allele likelihood subset.

[0017] FIG. 5A to FIG. 5B An example of an accelerated genotype imputation system according to one or more embodiments of the present disclosure is to determine a running sum of intermediate allele probabilities for genomic regions that exhibit haplotype alleles for one or more haplotypes given a marker variant, and use the running sum as a running input to determine individual intermediate allele probabilities for genomic regions that exhibit haplotype alleles given another marker variant.

[0018] Figure 6 An accelerated genotype imputation system according to one or more embodiments of the present disclosure is illustrated storing haplotype allele indicator data for a haplotype matrix on a memory device and accessing the stored haplotype allele indicator data to determine a value as part of a round across the haplotype matrix.

[0019] Figure 7 An accelerated computing engine of an accelerated genotype estimation system according to one or more embodiments of the present disclosure is illustrated.

[0020] Figure 8 The data flow engine of the accelerated genotype estimation system according to one or more embodiments of the present disclosure is illustrated to orchestrate the data input and output of the accelerated computing engine cluster and the on-board memory device of the configurable processor board.

[0021] Fig. 9 A schematic diagram of an accelerated computing engine including a core, peripheral interfaces, and other hardware according to one or more embodiments of the present disclosure is illustrated.

[0022] Fig.10 A series of actions for determining an intermediate allele likelihood for a genomic region including haplotype alleles by running a merge operation on a processor according to one or more embodiments of the present disclosure are illustrated.

[0023] Fig.11 A series of actions for determining and storing an intermediate allele likelihood subset as a hot start point corresponding to a marker variant group and temporarily generating an intermediate allele likelihood set for a marker variant set by using the intermediate allele likelihood subset according to one or more embodiments of the present disclosure are illustrated.

[0024] Fig.12Illustrated is a series of actions for determining a running sum of intermediate allele probabilities for genomic regions that exhibit haplotype alleles for one or more haplotypes given a marker variant, and using the running sum as a running input to determine individual intermediate allele probabilities for genomic regions that exhibit haplotype alleles given another marker variant, according to one or more embodiments of the present disclosure.

[0025] Fig.13 A block diagram of an example computing device is illustrated in accordance with one or more embodiments of the present disclosure. DETAILED DESCRIPTION

[0026] The present disclosure describes one or more embodiments of an accelerated genotype imputation system, which can determine the intermediate allele probability of a genomic region exhibiting certain haplotype alleles by using merged calculations or efficient data transmission across dedicated hardware as part of a genotype imputation model. For example, the accelerated genotype imputation system can determine the intermediate allele probability of a genomic region including haplotype alleles given a specific marker variant and a haplotype from a haplotype reference group by running a single round of parallel multiplication operations on a processor, rather than multiple rounds of parallel multiplication operations. In some cases, the accelerated genotype imputation system (i) determines and stores a subset of the intermediate allele probability corresponding to the marker variant group, and (ii) generates a set of intermediate allele probabilities by using the intermediate allele probability subset as a hot start point for a complete round of intermediate allele probabilities. The accelerated genotype imputation system can perform (i) and (ii) without storing multiple complete intermediate allele probability sets on a processor chip during real-time processing. In certain further embodiments, the accelerated genotype imputation system determines a running sum of the median allele likelihoods for genomic regions that exhibit haplotype alleles for one or more haplotypes given a marker variant, and uses the running sum as a running input to determine the median allele likelihoods for genomic regions that exhibit haplotype alleles given another marker variant. By using such a running sum, the accelerated genotype imputation system avoids processors summing intermediate allele likelihoods for neighboring markers and / or generating idle waiting time periods for allele likelihoods that slow down existing sequencing systems.

[0027] As suggested above, the accelerated genotype inference system applies a genotype inference model (such as a model based on a hidden Markov model (HMM)) to the nucleotide reads of the genomic region from the genomic sample to determine the posterior genotype possibility and haplotype detection of the genomic region. To illustrate, in some embodiments, the accelerated genotype inference system determines that the genomic region includes a priori genotype possibility of a specific genotype (e.g., a reference allele or an alternative allele), wherein the genomic region corresponds to a variable position or coordinate of a haplotype reference group. Such prior genotype possibilities are based on the quality scores of the nucleotide reads and nucleotide reads from the genomic sample. The accelerated genotype inference system further deconvolutes the vector of the prior genotype possibility into two independent vectors of haplotype allele possibilities (or, simply put, haplotype possibilities), wherein each vector corresponds to a complementary haplotype in two complementary haplotypes. Based on the haplotype possibilities from the independent vectors, the accelerated genotype inference system uses the haploid version of the HMM to infer two target haplotypes as haplotype detections. The accelerated genotype imputation system further determines (and updates) the phase of the two imputed haplotypes. In some embodiments, the accelerated genotype imputation system uses a genotype likelihood imputation and phasing method (GLIMPSE) as a genotype imputation model, as described in Simone Rubinacci et al., "Efficient Phasing and Imputation of Low-coverage Sequencing Data Using Large Reference Panels", 53 Nature Genetics 120-126 (2021) (hereinafter, Rubinacci), which is hereby incorporated by reference in its entirety.

[0028] The disclosed accelerated genotype imputation system introduces and utilizes a unified computation or unique architecture to efficiently determine the intermediate allele likelihood of genomic regions exhibiting certain haplotype alleles as part of GLIMPSE or another genotype imputation model. The following paragraphs briefly introduce various embodiments of the accelerated genotype imputation system.

[0029] A. Single round parallel computing

[0030] As suggested above, the accelerated genotype inference system determines the intermediate allele possibility of the genomic region including the haplotype allele by running a single round of parallel multiplication operations for a given marker variant and haplotype. In order to perform such an operation, in some embodiments, the accelerated genotype inference system identifies the haplotype reference group of the genomic region of the genomic sample as a part of the genotype inference model. The accelerated genotype inference system also accesses the first transition perception allele possibility factor (e.g., Q[m][Allele]*P1[m]) corresponding to the haplotype allele from the haplotype reference group and the second transition perception allele possibility factor (e.g., Q[m][Allele]*P0[m]) corresponding to the haplotype allele. By combining the first allele probability factor and the neighboring marker intermediate allele probability of the genomic region including the haplotype allele under the given neighboring marker variant (e.g., A'[m-1][k]), the accelerated genotype imputation system can perform a single round of parallel multiplication and generate the neighboring marker transition factor-aware allele probability of the marker variant and haplotype (e.g., Q[m][Allele]*P1[m]*A'[m-1]). Based on the neighboring marker transition factor-aware allele probability and the second transition-aware allele probability factor, for the given marker variant and haplotype, the accelerated genotype imputation system further determines the intermediate allele probability of the genomic region including the haplotype allele.

[0031] By performing such a single round of parallel multiplication operations, the accelerated genotype imputation system speeds up the computer processing time to determine the intermediate allele possibility and the allele possibility, while the computer processing time of the existing sequencing system is slower. As mentioned above, some existing sequencing systems running a single thread on a central processing unit (CPU) consume an average of about 17.5 hours to phase and impute the haplotype allele possibility corresponding to a single marker allele in a genomic region, wherein the HMM calculation time on a single CPU thread may consume about 10 hours. As further explained below, the several hours of long computer processing time partly comes from the sequencing system, which performs 3 multiplication operations to determine the intermediate allele possibility of each given marker variant and haplotype, and performs 3,000 multiplication operations for each haplotype of a haplotype reference group (e.g., organized in rows).

[0032] Compared to existing sequencing systems, in some embodiments, the disclosed accelerated genotype imputation system performs a single round of parallel multiplication operations to determine the intermediate allele likelihood for each given pair of marker variants and haplotypes, and due to such merged multiplication operations, approximately 1,000 multiplication operations are performed for each haplotype of a haplotype reference group (e.g., organized in rows). Together with other merge operations or other embodiments, the accelerated genotype imputation system can reduce the computer processing time for a single processor thread to perform approximately 40,000 HMM calculation tasks from approximately 10 hours or more (e.g., 600 minutes to 640 minutes) to approximately 60 seconds, thereby speeding up the processing time by 600 times.

[0033] B. Hot start intermediate allele likelihood subset

[0034] As further noted above, in some cases, the accelerated genotype inference system determines and stores a subset of the intermediate allele possibilities corresponding to the marker variant group, and temporarily generates a set of intermediate allele possibilities by using the subset of the intermediate allele possibilities as a hot start point for determining a complete round of intermediate allele possibilities. In order to determine and utilize such hot start possibilities, in some embodiments, the accelerated genotype inference system determines the first round of intermediate allele possibilities from the genomic region of a genomic sample in the case of a given marker variant set, and the genomic sample includes the haplotype alleles corresponding to the haplotype set. The accelerated genotype inference system also stores the first round of intermediate allele possibilities of the marker variant subset corresponding to the marker variant group on a dynamic random access memory (DRAM) or other storage device. Subsequently, the accelerated genotype inference system uses the stored subset of the first round of intermediate allele possibilities to initialize the allele possibilities at the marker variant group to determine, thereby regenerating the first round of intermediate allele possibilities. The accelerated genotype imputation system also uses the stored subset of the first round intermediate allele probabilities to determine a second round intermediate allele probabilities for a genomic region including haplotype alleles corresponding to the marker variant group and the haplotype group. Based on the regenerated first round intermediate allele probabilities and the second round intermediate allele probabilities, the accelerated genotype imputation system generates allele probabilities for a genomic region including haplotype alleles.

[0035] By determining and using such intermediate allele possibility subsets as hot start points, the accelerated genotype imputation system intelligently and efficiently redistributes data between memory devices, reducing data storage and increasing on-chip bandwidth. As described above, some existing sequencing systems (such as GLIMPSE) that use HMM for genotype determination determine and save values ​​in a haplotype matrix of about 50 million units. The data of this haplotype matrix will be saturated or prove too much to be stored in the on-chip memory of other processors of a field programmable gate array (FPGA) or an existing sequencing system. In order to reduce and redistribute the data of such a huge haplotype matrix, in some embodiments, the accelerated genotype imputation system determines and stores a subset of intermediate allele possibilities corresponding to a marker variant group, and uses the intermediate allele possibility subset as a hot start point for determining a complete round of intermediate allele possibilities.

[0036] By determining and storing intermediate allele possibility subsets, the accelerated genotype imputation system can exponentially reduce and transfer data, depending on the size of the marker variant group or window. In some embodiments, for example, the accelerated genotype imputation system reduces the memory load by 100 times by determining and storing the intermediate allele possibility subset corresponding to each marker variant from a 100-count marker variant group, or reduces the memory load by 1,000 times by determining and storing the intermediate allele possibility subset corresponding to each marker variant from a 1,000-count marker variant group. As further explained below, in some embodiments, the size of the marker variant group controls the exponent of the accelerated genotype imputation system to reduce memory load and data transfer.

[0037] C. Run the sum of the likelihoods of the intermediate alleles of neighboring markers

[0038] In addition to the rounds of concurrent multiplication operations and hot-starting intermediate allele likelihood subsets, the accelerated genotype imputation system can also determine the running sum of the intermediate allele likelihoods of genomic regions that exhibit haplotype alleles for one or more haplotypes given a marker variant, and use the running sum as a running input to determine the individual intermediate allele likelihoods of genomic regions that exhibit haplotype alleles given another marker variant. In order to utilize such running sums, in some embodiments, the accelerated genotype imputation system identifies a haplotype reference group of genomic regions of a genomic sample as part of a genotype imputation model. The accelerated genotype imputation system further (i) determines the running sum of a first subset of intermediate allele likelihoods of genomic regions including a first type of haplotype alleles of one or more haplotypes from the haplotype reference group for adjacent marker variants, and (ii) determines the running sum of a second subset of intermediate adjacent allele likelihoods of genomic regions including a second type of haplotype alleles from one or more haplotypes for adjacent marker variants. Based on the running sum, for a marker variant, the accelerated genotype imputation system determines the sum of the intermediate allele likelihoods for the genomic region that includes the haplotype alleles from the haplotype reference set.

[0039] By determining and using such a running sum of intermediate allele probabilities, the accelerated genotype imputation system eliminates or reduces wait time periods for performing one or both of the following: summing the intermediate allele probabilities for adjacent markers and generating the allele probabilities. As described above, some existing sequencing systems sum the intermediate allele probabilities and determine the allele probabilities for one marker variant before determining the respective intermediate allele probabilities for another marker variant, thereby causing the processor of the existing sequencing system to wait through a certain wait time period to perform one or both of the following: summing the intermediate allele probabilities for adjacent markers and generating the allele probabilities for another marker variant.

[0040] Compared to existing sequencing systems, in some embodiments, the accelerated genotype imputation system determines the running sum of the intermediate allele possibilities of the genomic region showing haplotype alleles for one or more haplotypes given a marker variant as a running input, and there is no conventional waiting time for determining the individual intermediate allele possibilities of the genomic region showing haplotype alleles given another marker variant. In the absence of such a waiting time cycle for summing the intermediate allele possibilities of adjacent markers and generating allele possibilities, the accelerated genotype imputation system accelerates the determination of the haplotype allele possibilities for the genotype imputation model faster than the existing sequencing system. Together with other merge operations or other embodiments, the accelerated genotype imputation system can reduce the computer processing time of a single processor thread to perform approximately 40,000 HMM calculation tasks from approximately 10 hours or more (e.g., 600 minutes to 640 minutes) to approximately 60 seconds, thereby speeding up the processing time by 600 times.

[0041] D. Customized hardware architecture

[0042] In order to promote the calculation or data storage of one or more mergers, in some embodiments, the accelerated genotype inference system utilizes a custom architecture. For example, the accelerated genotype inference system can store the intermediate allele possibility subset on a dynamic random access memory (DRAM) or another memory device (and access it from it) to hot start a complete round of determining the intermediate allele possibility. As another example, the accelerated genotype inference system can use a data flow engine as a part of a configurable processor to (i) query and manage HMM calculation tasks for the corresponding accelerated computing engine cluster, and (ii) distribute input values ​​from the cluster to each accelerated computing engine to determine the intermediate allele possibility (or other HMM calculation tasks) of a column or matrix. In some cases, for example, the accelerated genotype inference system sends a corresponding input value set (e.g., allele possibility factor, conversion coefficient and haplotype allele value) to the corresponding accelerated computing engine from the data flow engine, and uses the corresponding accelerated computing engine to determine the corresponding intermediate allele possibility set corresponding to the corresponding subset of the marker variant and the corresponding subset of the haplotype based on the corresponding input value set.

[0043] As suggested above, the disclosed accelerated genotype imputation system uses a custom architecture that facilitates faster throughput of input and output values ​​of allele likelihoods in the genotype imputation model than the conventional and undifferentiated architecture of existing sequencing systems. For example, the accelerated genotype imputation system can use off-chip DRAM or other memory devices to store and quickly transfer intermediate allele likelihood subsets for hot starts, rather than slowing throughput by relying on on-chip memory to store values ​​of intermediate allele likelihoods. As described above, existing HMM-based genotype imputation can burden the bandwidth of high-speed buses (such as Peripheral Component Interconnect Express (PCIe)) or other interfaces with values ​​of 50 million cells for haplotype matrices (sometimes passing through 40,000 such matrices). In some embodiments, by storing and accessing haplotype allele indicator data for a haplotype matrix in on-chip DRAM (or other on-chip memory) from on-chip DRAM (or other on-chip memory), an accelerated genotype-inference system can generate allele probabilities using a hidden Markov haploid genotype inference model or a hidden Markov diploid genotype inference model with 4 or more gigabytes of PCIe bandwidth over existing sequencing systems.

[0044] In some embodiments, by using a data flow engine as part of a configurable processor and orchestrating the data flow to a cluster of accelerated computing engines, the accelerated genotype imputation system avoids waiting time periods and determines the allele likelihood of different haplotypes and marker variants in parallel. In fact, as further explained below, the data flow engine of the disclosed accelerated genotype imputation system can efficiently distribute input values ​​and output values ​​to different clusters of accelerated computing engines to determine the allele likelihood of equivalents from 6 trillion cells across multiple haplotype matrices in approximately 60 seconds.

[0045] As shown in the above discussion, the present disclosure uses a variety of terms to describe the features and advantages of the accelerated genotype inference system. As used herein, the term "genomic sample" refers to a target genome or a portion of a genome that undergoes sequencing. For example, a sample genome includes a nucleotide sequence (or a copy of such a separated or extracted sequence) separated or extracted from a sample organism. Specifically, a sample genome includes a full genome that is separated or extracted (in whole or in part) from a sample organism and is composed of nitrogen-containing heterocyclic bases. The sample genome may include fragments of deoxyribonucleic acid (DNA), ribonucleic acid (RNA), or other polymeric forms of nucleic acids or chimeric or hybrid forms of nucleic acids described below. In some cases, the sample genome is present in a sample prepared or separated by a kit and received by a sequencing device.

[0046] As used herein as well, the term "haplotype" refers to a nucleotide sequence present in an organism (or present in an organism from a population) and inherited from one or more ancestors. Specifically, a haplotype may include an allele or other nucleotide sequence present in the organism of a population and inherited from a single parent by these organisms together. In one or more embodiments, a haplotype includes a group of SNPs that tend to be inherited together on the same chromosome. As described below, in some cases, a haplotype from a haplotype reference group may be represented as "k", and a row of different haplotypes from a haplotype reference group may be represented as "k". Additionally, "estimated haplotypes" refer to estimates or statistical inferences of haplotypes present in a sample genome. For example, the inferred haplotype may be a statistically inferred haplotype based on the genomic coordinates or regions of the SNPs surrounding or flanking the genomic coordinates or regions. As described above, the inferred haplotype may include SNPs or other variant nucleotide base calls around the target genomic region, and the custom sequencing system infers haplotypes thereon.

[0047] Relatedly, the term "haplotype allele" refers to a version of a nucleobase or nucleotide sequence at a genomic coordinate or genomic region corresponding to a haplotype (such as a haplotype of a genomic region or a non-coding region of a coding gene). Specifically, a haplotype allele includes one of two or more versions of a nucleobase or nucleotide sequence at a genomic coordinate or region, and the two or more versions tend to be combined together as part of a haplotype and inherited. As part of a haplotype, in some cases, a combination of haplotype alleles can be inherited by an organism as part of a single gene or across multiple genes. In some cases, the present disclosure describes different types of haplotype alleles. For example, in some embodiments, a type of haplotype allele may refer to a sample reference haplotype allele, and another type of haplotype allele may refer to a sample replacement haplotype allele. Although the present disclosure sometimes describes a first type and a second type of haplotype allele corresponding to a specific haplotype, in some embodiments, a haplotype may include more than two types of haplotype alleles (e.g., a sample reference haplotype allele and a plurality of sample replacement haplotype alleles).

[0048] In some cases, haplotype or its composition haplotype allele are represented by haplotype reference group.As used herein, term " haplotype reference group " refers to the digital collection or database of the haplotype from genomic sample, and one or more ancestors or ancestor haplotypes have been determined for this genomic sample.In some cases, haplotype reference group includes the digital database of the haplotype from genomic sample, and this genomic sample represents the colony of organism (or common in the colony of organism) and has determined multiple ancestors or ancestor haplotypes for this genomic sample.In some cases, the accelerated genotype inference system uses the haplotype reference group developed by haplotype reference alliance (HRM), 1000 genome project or Illumina, Inc.

[0049] Relatedly, the term "genotype inference model" refers to an algorithm or model for evaluating the genotype of a genomic region based on sequencing data from a genomic sample and the haplotype corresponding to the corresponding genomic region. Specifically, the genotype inference model includes an algorithm or model based on a hidden Markov model (HMM) for evaluating the genotype and phased haplotype of a genomic region based on sequencing data from a genomic sample and the haplotype corresponding to the corresponding genomic region from a haplotype reference group. As indicated above, in some cases, the genotype inference model includes GLIMPSE. Alternatively, the genotype inference model includes fastPHASE, BEAGLE, MACH or IMPUTE.

[0050] As a part of inferring genotype, in some cases, the accelerated genotype inference system determines allele probability. As used herein, the term "allele probability" refers to the possibility that a genomic region exhibits or includes a haplotype allele corresponding to a haplotype. For example, in some embodiments, the allele probability includes that a genomic region of a genomic sample exhibits or includes a sample reference haplotype allele or a sample substitution haplotype allele of a specific haplotype from a haplotype reference group. As described below, in some cases, the allele probability can be expressed as (i) R0, representing the possibility that a genomic region of a genomic sample includes a sample reference haplotype allele of a specific haplotype, or (ii) R1 represents the possibility that a genomic region of a genomic sample includes a sample substitution haplotype allele of a specific haplotype. Therefore, in some cases, the allele probability represents the posterior genotype probability generated by the genotype inference model.

[0051] Relatedly, the term "intermediate allele likelihood" refers to a value representing a temporary or preliminary likelihood that a genomic region exhibits or includes a haplotype allele corresponding to a haplotype. For example, in some embodiments, the intermediate allele likelihood includes a value representing a temporary or preliminary likelihood that a genomic region of a genomic sample exhibits or includes a sample reference haplotype allele or a sample alternative haplotype allele for a specific haplotype from a haplotype reference group in the case of a given target marker variant. As further described below, in some cases, the intermediate allele likelihood may be represented as A[m][k] and referred to as an α value, or alternatively, represented as B[m][k] and referred to as a β value. Although the present disclosure primarily uses A[m][k] as an example symbol for intermediate allele likelihood in α rounds, the symbol B[m][k] may be used interchangeably for intermediate allele likelihood in β rounds.

[0052] Relatedly, the term "marker variant" refers to a variant at a polymorphic site in a population. Specifically, a marker variant includes an allele in two or more alleles present in a population at a polymorphic genomic coordinate or genomic region with a frequency greater than a threshold frequency (such as a population greater than 1%). In some cases, a marker variant includes a SNP present at a polymorphic genomic coordinate in a population. Additionally or alternatively, a marker variant may include an insertion or deletion (indel), a structural variant or other variant at a polymorphic site in a population. As further described below, in some cases, a marker variant or a target marker variant is expressed as m or [m]. On the contrary, the term "adjacent marker variant" refers to a marker variant sorted before or after a target marker variant according to a specific order. Specifically, adjacent marker variants include marker variants represented by adjacent columns, which are located one column before or one column after the target column of the target marker variant in the representation matrix. As further explained below, in some cases, adjacent marker variants are expressed as m-1 or [m-1] or m+1 or [m+1].

[0053] Relatedly, as used herein, the term "adjacent marker intermediate allele possibility" refers to the intermediate allele possibility of the marker variant adjacent to the target marker variant. Specifically, the adjacent marker intermediate allele possibility comprises the intermediate allele possibility of the marker variant represented by the adjacent column, which is positioned at a column before or after the target column of the target marker variant in the representation matrix. As further explained below, in some cases, the adjacent marker intermediate allele possibility is expressed as A[m-1][k].

[0054] As further used herein, the term "allele probability factor" refers to a factor or parameter corresponding to a haplotype allele and applied to a conversion coefficient in a function and / or other parameters in a function. Specifically, the allele probability factor comprises a factor or parameter that (i) corresponds to a sample reference haplotype allele or a sample alternative haplotype allele and marker variants, and (ii) is applied to other parameters in a conversion linear coefficient, a conversion constant coefficient and / or a function to determine the allele probability. As further explained below, in some cases, the allele probability factor is usually expressed as Q[m][Allele], the allele probability factor corresponding to the sample reference haplotype allele is expressed as Q0, and the allele probability factor corresponding to the sample alternative haplotype allele is expressed as Q1.

[0055] Relatedly, the term "transition coefficient" refers to a coefficient or parameter representing the probability of transition or change between marker variants. Specifically, the transition coefficient includes a coefficient or parameter representing the probability of transition between rows representing marker variants within a matrix. In some cases, the transition coefficient has several forms, including a transition linear coefficient and a transition constant coefficient. As described below, in some cases, the transition constant coefficient is represented as P0, and the transition linear coefficient is represented as P1.

[0056] In some cases, the accelerated genotype inference system combines (e.g., multiply, weighted sum) various factors or coefficients. For example, as used herein, the term "transition perception allele possibility factor" refers to the value of the combination representing the transition coefficient and the allele possibility factor. Specifically, the transition perception allele possibility factor includes the value representing the product of the transition coefficient and the allele possibility factor. As described below, in some cases, the transition perception allele possibility factor is generally expressed as Q[m][Allele]*P[m], the first transition perception allele possibility factor is expressed as Q[m][Allele]*P1[m], and the second transition perception possibility factor is expressed as Q[m][Allele]*P0[m].

[0057] As further used herein, term " contiguous marker conversion factor perception allele possibility " refers to the value of the combination of expression allele possibility factor, conversion coefficient and contiguous marker variant intermediate allele possibility.Particularly, contiguous marker conversion factor perception allele possibility comprises the value of the product of expression allele possibility factor, conversion linear coefficient and contiguous marker variant intermediate allele possibility.As described below, in some cases, contiguous marker conversion factor perception allele possibility is usually expressed as Q[m][Allele]*P1[m]*A'[m-1].

[0058] As further used herein, term " adjacent mark conversion perception allele possibility factor of summation " refers to the value of the combination of the summation of the intermediate allele possibility of expression allele possibility factor, conversion coefficient and adjacent mark variant.Particularly, the adjacent mark conversion perception allele possibility factor of summation comprises the value of the product of the summation of the intermediate allele possibility of expression allele possibility factor, conversion constant coefficient and adjacent mark variant.As described below, in some cases, the adjacent mark conversion perception allele possibility factor of summation is usually expressed as Q[m][Allele]*P0[m]*Sum'[m-1].

[0059] As indicated above, in some embodiments, the accelerated genotype inference system can temporarily generate a set of intermediate allele possibilities for multiple rounds by using an intermediate allele possibility subset as a hot start point through which the intermediate allele possibility is completely passed. As used herein, the term "round" refers to an operation sequence that determines the intermediate allele possibility corresponding to a haplotype from a haplotype reference group according to a specific direction. Specifically, a round includes a series of operations in the direction across a haplotype matrix to determine the intermediate allele possibilities corresponding to different combinations of marker variants and haplotypes from a haplotype reference group. For example, a round can be performed across a haplotype matrix in a forward direction or a reverse direction. In some cases, a round comprising an operation sequence from left to right of a haplotype matrix constitutes an α round, and a round comprising an operation sequence from right to left of a haplotype matrix constitutes a β round.

[0060] Relatedly, the phrase "round intermediate allele likelihood" refers to the set of intermediate allele likelihoods corresponding to the round. Specifically, the first round intermediate allele likelihood includes the set of intermediate allele likelihoods determined by performing the first round of operations in the first direction. In contrast, the second round intermediate allele likelihood set includes the set of intermediate allele likelihoods determined by performing the second round of operations in the second direction. For example, when the accelerated genotype inference system performs the first round in the backward direction across the haplotype matrix, the first round intermediate allele likelihood can be determined, and when the accelerated genotype inference system performs the second round in the forward direction across the haplotype matrix, the second round intermediate allele likelihood can be determined, and vice versa.

[0061] As indicated above, in some embodiments, the accelerated genotype inference system stores or accesses a subset of the first round or second round of intermediate allele possibilities corresponding to a subset of marker variants of a marker variant group. As used herein, the term "marker variant group" refers to a segment or window of a marker variant in a larger marker variant set. For example, a marker variant group may include multiple groups of 100, 1,000 or 5,000 marker variants in a set of 50,000 marker variants that are continuously ordered. Since the haplotype matrix can represent marker variant sets by column, wherein each individual column represents an individual marker variant, the marker variant group may also correspond to a row group. Therefore, a subset of the first round or second round of intermediate allele possibilities corresponding to a subset of marker variants may refer to a subset including an intermediate allele possibility for a marker variant from each marker variant group, such as 1 marker variant corresponding to every 100, 1,000 or 5,000 marker variants.

[0062] As further indicated above, in some embodiments, the accelerated genotype imputation system determines different running sums of different subsets of intermediate allele probabilities for genomic regions including different types of haplotype alleles. As used herein, the term "running sum of a subset of intermediate allele probabilities" refers to the sum value of one or more intermediate allele probabilities of a marker variant (e.g., a neighboring marker variant), and the sum value can be updated when additional intermediate allele probabilities are determined. Specifically, the running sum of a subset of intermediate allele probabilities includes the sum value of multiple intermediate allele probabilities of a genomic region that exhibits or includes a specific type of haplotype allele of one or more haplotypes from a haplotype reference group in the case of a given neighboring marker variant, wherein the sum value can be updated when additional intermediate allele probabilities corresponding to neighboring marker variants are determined. Thus, in some embodiments, the accelerated genotype imputation system (i) determines, for neighboring marker variants, a running sum of a first subset of intermediate allele probabilities for genomic regions that include a first type of haplotype alleles from one or more haplotypes from a haplotype reference set (e.g., sample reference haplotype alleles), and (ii) determines, for neighboring marker variants, a running sum of a second subset of intermediate neighboring allele probabilities for genomic regions that include a second type of haplotype alleles from one or more haplotypes (e.g., sample alternative haplotype alleles).

[0063] In addition, as used herein, the term "genomic coordinates" refers to a specific position or orientation of a nucleotide base in a genome (e.g., a genome or reference genome of an organism). In some cases, the genomic coordinates include an identifier of a specific chromosome of a genome and an identifier of the orientation of a nucleotide base in a specific chromosome. For example, one or more genomic coordinates may include a chromosome number, a name or other identifier (e.g., chr1 or chrX) and one or more specific positions, such as a numbered position after the chromosome identifier (e.g., chr1:1234570 or chr1:1234570-1234870). In addition, in some specific implementations, genomic coordinates refer to the source of a reference genome (e.g., mitochondrial DNA reference genome mt or SARS-CoV-2 reference genome of a SARS-CoV-2 virus) and the position of a nucleotide base in the source of a reference genome (e.g., mt:16568 or SARS-CoV-2:29001). In contrast, in some cases, genomic coordinates refer to the position of a nucleotide base in a reference genome, without reference to a chromosome or source (e.g., 29727).

[0064] In addition, as used herein, "genomic region" refers to the range of genomic coordinates. Like genomic coordinates, in certain embodiments, a genomic region can be identified by a chromosome identifier and one or more specific positions, such as the numbered positions after the chromosome identifier (e.g., chr1:1234570-1234870).

[0065] As used herein, for example, the term "configurable processor" refers to a circuit or chip that can be configured or customized to execute a specific application. For example, a configurable processor includes an integrated circuit chip designed to be configured or customized on-site by an end-user's computing device to execute a specific application. Configurable processors include, but are not limited to, ASICs, ASSPs, coarse-grained reconfigurable arrays (CGRAs), or FPGAs. In contrast, configurable processors do not include CPUs or GPUs. In some embodiments, the accelerated genotype inference system uses a configurable processor (e.g., FPGA) or a processor (e.g., CPU) to execute the various embodiments described herein.

[0066] As further used herein, the term "nucleobase call" (or simply "base call") refers to the determination or prediction of a specific nucleobase (or nucleobase pair) of a genomic coordinate of an oligonucleotide (e.g., a read) or a sample genome during a sequencing cycle. Specifically, a nucleobase call can indicate: (i) the determination or prediction of the type of nucleobase that has been incorporated into an oligonucleotide on a nucleotide sample slide (e.g., a read-based nucleobase call); or (ii) the determination or prediction of the type of nucleobase present at a genomic coordinate or region within a genome, including variant calls or non-variant calls in a digital output file. In some cases, for nucleotide fragment reads, a nucleobase call includes determining or predicting a nucleobase based on intensity values ​​generated by fluorescently labeled nucleotides of an oligonucleotide added to a nucleotide sample slide (e.g., in a cluster of a flow cell). Alternatively, a nucleobase call includes determining or predicting a nucleobase based on chromatographic peaks or current changes generated by nucleotides passing through a nanopore of a nucleotide sample slide. In contrast, based on the nucleotide fragment reads corresponding to the genomic coordinates, the nucleobase call may also include the nucleobase at the genomic coordinates of the sample genome of the final predicted variant call file (VCF) or other base call output file. Therefore, the nucleobase call may include base calls corresponding to the genomic coordinates and the reference genome, such as an indication of a variant or non-variant at a specific position corresponding to the reference genome. In fact, the nucleobase call may refer to a variant call, including but not limited to a single nucleotide variant (SNV), an insertion or deletion (indel), or a base call as part of a structural variant. As mentioned above, a single nucleobase call may be an adenine (A) call, a cytosine (C) call, a guanine (G) call, or a thymine (T) call.

[0067] As further used herein, the term "nucleotide sample slide" refers to a plate or slide comprising oligonucleotides for sequencing the nucleotide sequence from a genomic sample or other sample nucleic acid polymer. Specifically, a nucleotide sample slide can refer to a slide containing a fluid channel, through which reagents and buffers can travel as part of sequencing. For example, in one or more embodiments, the nucleotide sample slide includes a flow cell (e.g., a patterned flow cell or an unpatterned flow cell) comprising a small fluid channel and a short oligonucleotide complementary to a binding adapter sequence. As indicated above, a nucleotide sample slide can include a hole (e.g., a nanopore) containing an oligonucleotide cluster.

[0068] As mentioned above, the circulation cell or other nucleotide sample slide may (i) include a device with a cap, which extends above the reaction structure to form a flow channel connected to multiple reaction sites of the reaction structure therebetween, and may (ii) include a detection device configured to detect a specified reaction occurring at or near the reaction site. The circulation cell or other nucleotide sample slide may include a solid-state light detection or imaging device, such as a charge coupled device (CCD) or a complementary metal oxide semiconductor (CMOS) (light) detection device. As a specific example, the circulation cell may be configured to be fluidically coupled and electrically coupled to a box (with an integrated pump), which may be configured to be fluidically coupled and / or electrically coupled to a bioassay system. The box and / or the bioassay system may deliver the reaction solution to the reaction site of the circulation cell according to a predetermined scheme (e.g., sequencing by synthesis), and perform multiple imaging events. For example, the box and / or the bioassay system may guide one or more reaction solutions through the flow channel of the circulation cell, thereby flowing along the reaction site. At least one of the reaction solutions may include four types of nucleotides with the same or different fluorescent labels. The nucleotides may be bound to the reaction sites of the flow cell, such as to the corresponding oligonucleotides at the reaction sites. The box and / or the bioassay system then illuminates the reaction sites using an excitation light source (e.g., a solid-state light source, such as a light emitting diode (LED)). The excitation light may provide an emission signal (e.g., light of one or more wavelengths different from the excitation light and possibly different from each other) that may be detected by the light sensor of the flow cell.

[0069] As further used herein, the term "sequencing run" refers to an iterative process of determining the primary structure of a nucleotide sequence from a sample (e.g., a genomic sample) on a sequencing device. Specifically, a sequencing run includes cycles of sequencing chemistry and imaging performed by a sequencing device that incorporates nucleobases into growing oligonucleotides to determine nucleotide fragment reads from nucleotide sequences extracted from samples (or other sequences within library fragments) and inoculated across a nucleotide sample slide. In some cases, a sequencing run includes replicating nucleotide sequences from one or more genomic samples that are inoculated as clusters across a nucleotide sample slide (e.g., a flow cell). Upon completion of the sequencing run, the sequencing device may generate base call data in a file.

[0070] As indicated, the term "base call data" refers to data representing the nucleobase calls of nucleotide fragment reads and / or corresponding sequencing metrics. For example, the base call data includes text data representing the nucleobase calls of nucleotide fragment reads as text (e.g., A, C, G, T) and corresponding base call quality metrics, depth metrics, and / or other sequencing metrics. In some cases, the base call data is formatted in a text file (such as a binary base call (BCL) sequence file or a fast overall quality (FASTQ) file).

[0071] As further used herein, the term "nucleotide fragment read" (or simply "read") refers to a sequence of one or more nucleobases (or nucleobase pairs) inferred from all or part of a sample nucleotide sequence (e.g., a sample genomic sequence, cDNA). Specifically, a nucleotide fragment read includes a sequence determined or predicted based on nucleobase calls of a nucleotide sequence (or a monoclonal nucleotide sequence set) corresponding to a sample library fragment of a genomic sample. For example, in some cases, a sequencing device determines a nucleotide fragment read by generating nucleobase calls of nucleobases that pass through a nanopore of a nucleotide sample slide, by determining via fluorescent labeling, or by determining based on clusters in a circulation pool.

[0072] The following paragraphs describe the accelerated genotype estimation system with respect to illustrative drawings that depict exemplary embodiments and specific implementations. For example, Figure 1 A schematic diagram of a computing system 100 in which an accelerated genotype estimation system 106 operates according to one or more embodiments is illustrated. As illustrated, the computing system 100 includes a sequencing device 102 connected to a local device 110 (e.g., a local server device), one or more server devices 120, and a client device 116. Figure 1 As shown, sequencing device 102, local device 110, server device 120, and client device 116 can communicate with each other via network 122. Network 122 includes any suitable network over which computing devices can communicate. Fig.13 The example network is discussed in more detail. Figure 1 An embodiment of the accelerated genotype imputation system 106 is shown, but the present disclosure describes the following alternative embodiments and configurations.

[0073] like Figure 1As indicated, the sequencing device 102 includes a computing device and a sequencing device system 104 for sequencing a genomic sample or other nucleic acid polymer. In some embodiments, the sequencing device 102 analyzes nucleotide fragments or oligonucleotides extracted from the genomic sample by executing the sequencing device system 104 on a processor 108 (e.g., a configurable processor) to generate nucleotide fragment reads or other data directly or indirectly on the sequencing device 102 using computer-implemented methods and systems. More specifically, the sequencing device 102 receives a nucleotide sample slide (e.g., a flow cell) including nucleotide fragments extracted from the sample, and further copies and determines the nucleobase sequence of such extracted nucleotide fragments.

[0074] In one or more embodiments, the sequencing device 102 utilizes SBS to sequence the nucleotide fragments into nucleotide fragment reads and determine the nucleobase calls of the nucleotide fragment reads. In addition or as an alternative to communicating across the network 122, in some embodiments, the sequencing device 102 bypasses the network 122 and communicates directly with the local device 110 or the client device 116. By executing the sequencing device system 104, the sequencing device 102 may also store the nucleobase calls as part of the base call data formatted as a BCL file and send the BCL file to the local device 110 and / or the server device 120.

[0075] like Figure 1 As further indicated, the local device 110 is located at or near the same physical location as the sequencing device 102. In fact, in some embodiments, the local device 110 and the sequencing device 102 are integrated into the same computing device. The local device 110 can operate the sequencing system 112 to generate, receive, analyze, store, and transmit digital data, such as by receiving base call data or determining variant calls based on analyzing such base call data. Figure 1 As shown, the sequencing device 102 can send (and the local device 110 can receive) base call data generated during a sequencing run of the sequencing device 102. By executing software in the form of a sequencing system 112, the local device 110 can align the nucleotide fragment reads with the reference genome and determine genetic variations based on the aligned nucleotide fragment reads. The local device 110 can also communicate with a client device 116. Specifically, the local device 110 can send data to the client device 116, which includes a variant call file (VCF) or other information indicating a nuclear base call, sequencing metrics, error data, or other metrics.

[0076] As indicated above, as part of the local device 110, the accelerated genotype imputation system 106 can determine the intermediate allele likelihood of a genomic region exhibiting certain haplotype alleles by using one or both of merged calculations and data exchange across dedicated hardware as part of a genotype imputation model. For example, the accelerated genotype imputation system 106 can determine the intermediate allele likelihood of a genomic region including a haplotype allele given a specific marker variant and a haplotype from a haplotype reference group by running a single round of parallel multiplication operations on the processor 114. In some specific implementations, the processor 114 is a configurable processor. In some cases, the accelerated genotype imputation system 106 (i) determines and stores a subset of the intermediate allele likelihood corresponding to the marker variant group, and (ii) temporarily generates a set of intermediate allele likelihoods for multiple rounds by using the intermediate allele likelihood subset as a hot start point for a complete round of intermediate allele likelihoods. In a further embodiment, the accelerated genotype imputation system 106 determines a running sum of the median allele likelihood for genomic regions that exhibit haplotype alleles for one or more haplotypes given one marker variant as a running input for determining the median allele likelihood for genomic regions that exhibit haplotype alleles given another marker variant.

[0077] like Figure 1 As further shown, the server device 120 is located remotely from the local device 110 and the sequencing device 102. Similar to the local device 110, in some embodiments, the server device 120 includes a version of the sequencing system 112. Thus, the server device 120 can generate, receive, analyze, store, and transmit digital data, such as by receiving base call data or determining variant calls based on analyzing such base call data. Thus, the sequencing device 102 can send (and the server device 120 can receive) base call data from the sequencing device 102. The server device 120 can also communicate with the client device 116. Specifically, the server device 120 can send data to the client device 116, which includes VCF or other sequencing related information.

[0078] In some embodiments, server device 120 comprises a distributed server collection, where server device 120 comprises many server devices distributed across network 122 and located in the same or different physical locations. In addition, server device 120 may comprise a content server, an application server, a communication server, a web hosting server, or another type of server.

[0079] like Figure 1As further illustrated and indicated, by generating a sequencing application 118, the client device 116 may generate, store, receive, and send digital data. Specifically, the client device 116 may receive sequencing data from the local device 110, or receive a call file (e.g., a BCL) and sequencing metrics from the sequencing device 102. In addition, the client device 116 may communicate with the local device 110 or the server device 120 to receive a VCF that includes nucleobase calls and / or other metrics, such as base call quality metrics or pass filter metrics. The client device 116 may accordingly present or display information related to variant calls or other nucleobase calls to a user associated with the client device 116 within a graphical user interface of the sequencing application 118. For example, the client device 116 may present variant calls and / or sequencing metrics for a sequenced genomic sample within a graphical user interface of the sequencing application 118.

[0080] although Figure 1 Client device 116 is depicted as a desktop computer or laptop computer, but client device 116 may include various types of client devices. For example, in some embodiments, client device 116 includes a non-mobile device, such as a desktop computer or server, or other types of client devices. In still other embodiments, client device 116 includes a mobile device, such as a laptop, tablet computer, mobile phone, or smartphone. Fig.13 Additional details regarding client device 116 are discussed.

[0081] like Figure 1 As further illustrated in FIG. 1 , the client device 116 includes a sequencing application 118. The sequencing application 118 may be a web application or a native application (e.g., a mobile application, a desktop application) stored and executed on the client device 116. The sequencing application 118 may include instructions that, when executed, cause the client device 116 to receive data from the accelerated genotype imputation system 106 and present base call data or data from the VCF for display at the client device 116. In addition, the sequencing application 118 may direct the client device 116 to display a summary for multiple sequencing runs.

[0082] like Figure 1As further illustrated in , a version of the accelerated genotype imputation system 106 may be located (e.g., completely or partially) on the local device 110 and implemented on the local device. In yet other embodiments, the accelerated genotype imputation system 106 is implemented by one or more other components of the computing system 100 (such as a server device 120). Specifically, the accelerated genotype imputation system 106 may be implemented in a variety of different ways across the sequencing device 102, the local device 110, the server device 120, and the client device 116. For example, the accelerated genotype imputation system 106 may be downloaded from the server device 120 to the accelerated genotype imputation system 106 and / or the local device 110, wherein all or part of the functionality of the accelerated genotype imputation system 106 is executed at each corresponding device within the computing system 100.

[0083] As suggested above, in some embodiments, the accelerated genotype imputation system 106 applies a genotype imputation model (such as a genotype imputation model based on a hidden Markov model (HMM)) to the nucleotide fragment reads corresponding to the genomic region of the genomic sample. By applying the genotype imputation model, the accelerated genotype imputation system 106 can determine the posterior genotype probability and haplotype call of the genomic region. According to one or more embodiments, Figure 2A The accelerated genotype imputation system 106 is illustrated to apply GLIMPSE as a genotype imputation model to determine the posterior genotype likelihood of genomic regions of multiple genomic samples. As part of using HMM to impute haplotypes, the accelerated genotype imputation system 106 uses the haplotype matrix 220 to determine the haplotype allele likelihood corresponding to the genomic region. According to one or more embodiments, Figure 2B A more detailed description of the accelerated genotype imputation system 106 utilizing the haplotype matrix 220 to determine such haplotype allele likelihoods is shown.

[0084] like Figure 2A As shown, for example, the accelerated genotype estimation system 106 determines a priori genotype likelihood 204 that a genomic region 200 from a plurality of genomic samples exhibits a particular genotype (e.g., a reference allele or an alternative allele). Figure 2A As suggested, in some cases, the genomic region 200 corresponds to a substantially identical set of genomic coordinates (relative to a reference genome) for a plurality of genomic samples. As indicated by the nucleotide fragment reads 202, the genomic region 200 exhibits low coverage (e.g., ≤8X read coverage). In some embodiments, the accelerated genotype imputation system 106 uses a probabilistic call generation model (e.g., a variant caller from DRAGEN) to determine a priori genotype likelihood 204 based on (i) nucleotide fragment reads 202 from a plurality of genomic samples and (i) quality scores of base calls of the nucleotide fragment reads 202.

[0085] like Figure 2A As further indicated, the genomic region 200 corresponds to a variable position (or variable genomic coordinate) of a haplotype reference set 206. The accelerated genotype imputation system 106 further deconvolutes the vector of a priori genotype probabilities 204 into two independent vectors of haplotype allele probabilities (or, simply, haplotype probabilities), wherein each vector corresponds to one of two complementary haplotypes. In some such embodiments, the accelerated genotype imputation system 106 inputs the a priori genotype probabilities 204 in vector form as part of an input matrix.

[0086] Based on the haplotype probabilities from the independent vectors, in some implementations, the accelerated genotype imputation system 106 imputes two target haplotypes as haplotype calls using the haploid version of the HMM in an iterative process. Figure 2A As shown, for example, the accelerated genotype imputation system 106 selects a haplotype 210 based on a haplotype reference set 206 and a target haplotype 208 estimated for each genomic sample. After selecting a haplotype for a given genomic sample, the accelerated genotype imputation system 106 stores the reference and target versions of the selected haplotype as a positional Burroughs-Wheeler transform (PBWT) 212.

[0087] like Figure 2A As further shown, in some embodiments, the accelerated genotype imputation system 106 samples haplotypes 214 in the PBWT 212 format by executing a linear time sampling algorithm based on a haplotype estimation version of the HMM developed by NaLi and Matthew Stephens in "Modeling Linkage Disequilibrium and Identifying Recombination Hotspots Using Single-Nucleotide Polymorphism Data", 165 Genetics 2213-2233 (2003), which is hereby incorporated by reference in its entirety. By executing the linear time sampling algorithm as part of the sampler iteration, the accelerated genotype imputation system 106 further determines (and updates) the phase of two imputed haplotypes of the genomic region 200 of the particular genomic sample.

[0088] like Figure 2AAs further shown, based on the imputed and phased haplotypes, the accelerated genotype imputation system 106 determines a posterior genotype likelihood 216 that a genomic region 200 of a plurality of genomic samples exhibits a particular genotype (e.g., a reference allele or an alternative allele). The accelerated genotype imputation system 106 further determines a haplotype call 218 for the genomic region of each genomic sample in the plurality of genomic samples. As indicated above, in some embodiments, the accelerated genotype imputation system 106 uses a modified version of GLIMPSE developed by Rubinacci as a genotype imputation model.

[0089] As part of selecting haplotypes 210 and sampling haplotypes 214, the accelerated genotype imputation system 106 may perform sampler iterations across genomic samples using the haplotype matrix 220. As further explained below and as described in Figure 2B Further depicted, the accelerated genotype imputation system 106 can determine the intermediate allele possibility of the genomic region including the haplotype allele in the forward direction and the reverse direction across the haplotype matrix 220. In the haplotype matrix 220, each column represents a marker variant, and each row represents a haplotype from a haplotype reference group 206. The accelerated genotype imputation system 106 further determines the sum of the intermediate allele possibilities of each column representing the marker variant. In some cases, based on the adjacent marker intermediate allele possibilities of the summation of each column, the accelerated genotype imputation system 106 determines the allele possibilities of the corresponding marker variant and the haplotype. Such allele possibilities represent examples or implementations of the posterior genotype possibilities 216.

[0090] like Figure 2B As shown, for example, the accelerated genotype estimation system 106 uses the input haplotype matrix 220a to input various values. Figure 2B As shown, the input haplotype matrix 220a and the updated haplotype matrix 220b are organized by "K" rows representing haplotypes from the haplotype reference group 206 and "M" columns representing marker variants (e.g., SNPs or other variants). Therefore, each row represents a haplotype "k", and each column represents a marker variant "m". In some embodiments, the input haplotype matrix 220a and the updated haplotype matrix 220b each include about 1,000 rows representing about 1,000 haplotypes from the haplotype reference group 206 and about 50,000 columns representing about 50,000 marker variants. Therefore, the input haplotype matrix 220a includes approximately 50 million cells. However, other suitable sizes of more or fewer columns and rows may also be used.

[0091] like Figure 2BAs further shown, in some embodiments, the accelerated genotype imputation system 106 inputs the values ​​of the transition coefficients (e.g., P0 and P1) and the allele likelihood factors (e.g., Q0 and Q1) into each cell of the input haplotype matrix 220a. For example, the accelerated genotype imputation system 106 inputs a specific transition linear coefficient (e.g., P1) and a specific transition constant coefficient (e.g., P0) into each cell, wherein the transition coefficient generally represents the probability of transitioning between haplotypes represented by adjacent rows. In addition, the accelerated genotype imputation system 106 inputs into each cell a specific allele likelihood factor (e.g., Q0) of a first type of haplotype allele of a specific haplotype represented by a row, and inputs a specific allele likelihood factor (e.g., Q1) of a second type of haplotype allele of a specific haplotype represented by the row. As described above, in some embodiments, one allele likelihood factor (e.g., Q0) corresponds to a sample reference haplotype allele for a particular haplotype represented by a row, and another allele likelihood factor (e.g., Q1) corresponds to a sample alternative haplotype for the particular haplotype.

[0092] In addition to entering the transition coefficient and allele likelihood factor, such as Figure 2B As further shown, in certain embodiments, the accelerated genotype inference system 106 inputs a value representing a haplotype allele (S bits) into each cell of the input haplotype matrix 220a. Specifically, the accelerated genotype inference system 106 may input a value (or bit) 0, indicating a sample reference haplotype allele for a particular haplotype represented by a row. Conversely, the accelerated genotype inference system 106 may input a value (or bit) 1, indicating a sample alternative haplotype allele for a particular haplotype represented by a row. For simplicity, the present disclosure refers to such input values ​​representing haplotype alleles as haplotype allele indicator data of a haplotype matrix, as described below with respect to Figure 6 Further described.

[0093] After inputting the values ​​of the transition coefficient, the allele probability factor, and the haplotype allele index, in some embodiments, the accelerated genotype inference system 106 determines the intermediate allele probability in each cell based on the input values. For example, in some embodiments, the accelerated genotype inference system 106 performs α rounds and β rounds across the cells of the input haplotype matrix 220a to determine the intermediate allele probability represented by the darker shading in the updated haplotype matrix 220b. In fact, in some embodiments, the α value represents the intermediate allele probability (e.g., A[m][k]) determined during the α round, and the β value represents the intermediate allele probability (e.g., A[m][k]) determined during the β round. As further described below, in some embodiments, the accelerated genotype inference system 106 performs two β rounds (including sacrificial β rounds) as part of the HMM calculation task.

[0094] In order to determine the intermediate allele probability (e.g., A[m][k]) of the target unit, in some embodiments, the accelerated genotype inference system 106 determines the first product of the conversion linear coefficient (e.g., P1[m]) of the target marker variant, the normalized value (e.g., Norm[m-1]) of the column representing the adjacent marker variant, and the adjacent marker intermediate allele probability (e.g., A[m-1][k]) of the adjacent marker variant. The normalized value (e.g., represented by the column) of a given marker variant can be any value that is convenient for preventing the per-unit value from overflowing the digital representation of the sum of the intermediate allele probability value or the intermediate allele probability value. The accelerated genotype inference system 106 also determines the second product of the conversion constant coefficient (e.g., P0[m]), the normalized value (e.g., Norm[m-1]) of the column representing the adjacent marker variant, and the adjacent marker intermediate allele probability of the adjacent marker variant (e.g., Sum[m-1]). The accelerated genotype imputation system 106 also multiplies the sum of the first product and the second product by an allele likelihood factor (eg, Q[m][Allele]) to determine the median allele likelihood of the target unit.

[0095] As described above, such an allele likelihood factor may constitute an allele likelihood factor (e.g., Q0) of a sample reference haplotype allele corresponding to a particular haplotype represented by a row or another allele likelihood factor (e.g., Q1) of a sample alternative haplotype corresponding to a particular haplotype. However, as described below, the accelerated genotype imputation system 106 may also perform an improved method of determining such intermediate allele likelihoods.

[0096] like Figure 2BAs further shown, in some embodiments, the accelerated genotype imputation system 106 determines the sum of the alpha values ​​of the marker variants and the sum of the beta values ​​of the marker variants for each column. Specifically, in some embodiments, the accelerated genotype imputation system 106 determines (i) the sum of the median allele likelihoods of the columns represented by the marker variants in one round and (ii) the sum of the median allele likelihoods of the columns represented by the marker variants in another round.

[0097] exist Figure 2B In some embodiments shown, based on the summed intermediate allele likelihood of each marker variant represented by the column, the accelerated genotype inference system 106 further determines a pair of allele probabilities (e.g., R0 and R1) for each marker variant. For example, in certain specific implementations, the accelerated genotype inference system 106 determines that the genomic region includes a first allele likelihood (e.g., R0) of the sample reference haplotype allele corresponding to each haplotype represented by each row. Similarly, the accelerated genotype inference system 106 determines that the genomic region includes a second allele likelihood (e.g., R1) of the sample alternative haplotype allele corresponding to each haplotype represented by each row.

[0098] As described above, in some cases, the accelerated genotype imputation system 106 accelerates the determination of intermediate allele likelihood by performing a single round of parallel multiplication operations on a given target marker variant and a haplotype from a haplotype reference set. According to one or more embodiments, Figure 3A The accelerated genotype imputation system 106 is depicted running a single round of parallel multiplication operations to determine an intermediate allelic likelihood for a genomic region including a haplotype allele given a target cell representing a marker variant and a target haplotype from a haplotype reference set. Figure 3B Depicted is a comparison of the accelerated genotype imputation system 106 determining such intermediate allele likelihoods for a target cell using (i) three rounds of parallel multiplication operations or (ii) one round of parallel multiplication operations. By predetermining the transition-aware allele likelihood factor before the processor determines the intermediate allele likelihood for a target marker variant, the accelerated genotype imputation system 106 condenses and speeds up the processing load from three rounds of parallel multiplication operations for a target cell to one round of parallel multiplication operations.

[0099] like Figure 3AAs shown, for example, the accelerated genotype inference system 106 identifies a haplotype reference group 304 and a transition-aware allele likelihood factor corresponding to a genomic region of one or more genomic samples from a storage device 302 to execute a genotype inference model. Specifically, in some embodiments, the accelerated genotype inference system 106 identifies a haplotype reference group 304 stored on a dynamic random access memory (DRAM), a dynamic random access memory (SRAM), or a cache memory device. In addition, the accelerated genotype inference system 106 identifies a first transition-aware allele likelihood factor 306a and a second transition-aware allele likelihood factor 306b while executing an alpha round or a beta round of a haplotype matrix 308. In some cases, the accelerated genotype inference system 106 identifies a first transition-aware allele likelihood factor 306a and a second transition-aware allele likelihood factor 306b when reaching a target unit 300 representing a combination of a target marker variant and a haplotype during a round of the haplotype matrix 308.

[0100] To avoid determining the first transition-aware allele likelihood factor 306a and the second transition-aware allele likelihood factor 306b during the round, in some embodiments, the accelerated genotype imputation system 106 predetermines the first transition-aware allele likelihood factor 306a and the second transition-aware allele likelihood factor 306b before determining the intermediate allele likelihood of the columns representing the target marker variants within the haplotype matrix 308. To predetermine the first transition-aware allele likelihood factor 306a, in some embodiments, the accelerated genotype imputation system 106 combines (e.g., multiplies, weighted sums) the allele likelihood factors for the haplotype alleles and the transition constant coefficients for transitioning between haplotypes from the haplotype reference group 304. Similarly, to predetermine the second transition-aware allele likelihood factor 306b, the accelerated genotype imputation system 106 combines (e.g., multiplies, weighted sums) the allele likelihood factors and the transition linear coefficients for transitioning between haplotypes from the haplotype reference group 304.

[0101] The accelerated genotype imputation system 106 may generate a predetermined version of the first transition-aware allele likelihood factor 306a and the second transition-aware allele likelihood factor 306b because the input values ​​are available before the round across the haplotype matrix 308 or at least before the intermediate allele likelihood of the target marker variant is determined. Since the accelerated genotype imputation system 106 can access (and identify) the allele likelihood factors and transition coefficients of the columns representing the target marker variants before determining the intermediate allele likelihood of the target marker variants, in some specific implementations, the accelerated genotype imputation system 106 generates a predetermined version of the first transition-aware allele likelihood factor 306a and the second transition-aware allele likelihood factor 306b. Therefore, in some embodiments, the accelerated genotype imputation system 106 predetermines the first transition-aware allele likelihood factor 306a and the second transition-aware allele likelihood factor 306b before determining one or more intermediate allele likelihoods corresponding to the marker variants as part of the round of the haplotype matrix 308.

[0102] As part of performing a round of determining intermediate allele probabilities, in some cases, the accelerated genotype imputation system 106 determines and accesses values ​​as part of a round across the haplotype matrix 308. To determine the intermediate allele probability 316 for the target unit 300, in some embodiments, the accelerated genotype imputation system 106 identifies neighboring marker intermediate allele probabilities 310 for marker variants that are adjacent to the target marker variant from the haplotype matrix 308. In the haplotype matrix 308, the neighboring columns represent neighboring marker variants that are adjacent to the target column representing the target marker variant. As part of the round across the haplotype matrix 308, in some embodiments, the accelerated genotype imputation system 106 determines the neighboring marker intermediate allele probabilities 310 for the combination of the neighboring marker variant and the target haplotype from the haplotype reference group 304 before determining the intermediate allele probability 316.

[0103] After identifying the relevant input values ​​for the multiplication operation, such as Figure 3AAs further shown, the accelerated genotype imputation system 106 combines the neighboring marker intermediate allele likelihood 310 and the first transition-aware allele likelihood factor 306a. Specifically, in some embodiments, the accelerated genotype imputation system 106 multiplies the neighboring marker intermediate allele likelihood 310 and the first transition-aware allele likelihood factor 306a during a round of the haplotype matrix 308. Because the accelerated genotype imputation system 106 determines both the neighboring marker intermediate allele likelihood 310 and the first transition-aware allele likelihood factor 306a before passing through the unit representing the target marker variant and the target haplotype, the accelerated genotype imputation system 106 can use this single round of parallel multiplication operations as part of determining the intermediate allele likelihood 316 of the target unit 300. Figure 3A As shown, the accelerated genotype imputation system 106 generates a neighboring marker conversion factor-aware allele likelihood 314 based on a combination of the neighboring marker intermediate allele likelihood 310 and the first conversion-aware allele likelihood factor 306a.

[0104] As further suggested above, in some embodiments, the accelerated genotype imputation system 106 determines an intermediate allele probability 316 for a genomic region including haplotype alleles based on the neighboring marker transition factor-aware allele probability 314 and the second transition-aware allele probability factor 306b. For example, in some embodiments, the accelerated genotype imputation system 106 determines the sum of the neighboring marker transition factor-aware allele probability 314 and the second transition-aware allele probability factor 306b to determine the intermediate allele probability 316. As further explained below, in certain implementations, the accelerated genotype imputation system 106 determines the intermediate allele probability 316 by determining the sum of the product of (i) the neighboring marker transition factor-aware allele probability 314 and (ii) the second transition-aware allele probability factor 306b and the summed neighboring marker intermediate allele probability 312 for neighboring marker variants.

[0105] As indicated above, the accelerated genotype imputation system 106 can reduce computer processing from three multiplication operations to one multiplication operation to determine the intermediate allele likelihood of a target unit. According to one or more embodiments, Figure 3B The accelerated genotype imputation system 106 is depicted using a configurable processor to execute a concatenated model 318 and a single multiplicative model 320 for determining intermediate allele likelihoods for target cells representing combinations of target marker variants and haplotypes within a haplotype matrix.

[0106] like Figure 3BAs shown, when using the multiplication model 318, the accelerated genotype estimation system 106 performs multiplication operations 334a, 334b, and 334c as part of determining the intermediate allele likelihood 332a of the target unit. Figure 3B The order shown briefly summarizes multiplication 334a, 334b ​​and 334c, but any order can be used. First, the accelerated genotype inference system 106 performs multiplication 334a by multiplying the transition constant coefficient 322 (e.g., P0) of the column representing the target marker variant with the addition adjacent marker intermediate allele possibility 324 (e.g., Sum[m-1]) of the adjacent marker variant. In some cases, the adjacent marker intermediate allele possibility 324 of summation is normalized (e.g., Norm[m-1]*Sum[m-1]). For simplicity, the present invention uses apostrophes as abbreviation to indicate normalized values ​​(e.g., Sum'[m-1]).

[0107] Second, the accelerated genotype imputation system 106 performs a multiplication operation 334b ​​by multiplying the linear coefficient 326 (e.g., P1) of the column representing the target marker variant with the neighboring marker median allele likelihood 328a (e.g., A[m-1][k]) of the neighboring marker variant. In some cases, the neighboring marker median allele likelihood 328a is normalized (e.g., Norm[m-1]*A[m-1][k]). Figure 3B As further shown, the accelerated genotype estimation system 106 performs a summation operation 340a by summing the product of (i) the transition constant coefficient 322 (P0) and the summed adjacent marker intermediate allele probability 324 (e.g., Norm[m-1]*Sum[m-1]) and (i) the transition linear coefficient 326 (P0) and the adjacent marker intermediate allele probability 328a (e.g., Norm[m-1]*A[m-1][k]).

[0108] Third, the accelerated genotype imputation system 106 performs a multiplication operation 334c by multiplying the allele likelihood factor 330a (e.g., Q0 or Q1) of the column representing the target marker variant with the sum product. As suggested above, the allele likelihood factor 330a may constitute an allele likelihood factor (e.g., Q0) corresponding to the sample reference haplotype allele of the target haplotype represented by the row or another allele likelihood factor (e.g., Q1) corresponding to the sample alternative haplotype of the target haplotype. Based on multiplying the allele likelihood factor 330a (e.g., Q0 or Q1) with the sum product (P1[m]*Norm[m-1]*A[m-1][k]+P0[m]*Norm[m-1]*Sum[m-1]), the accelerated genotype imputation system 106 uses the multi-multiplication model 318 to determine the intermediate allele likelihood 332a (e.g., A[m][k]).

[0109] When using the multiplication model 318, in some embodiments, the accelerated genotype imputation system 106 determines the intermediate allele possibility of the target unit during the alpha round and the beta round. Therefore, the value corresponding to the adjacent marker variant (m-1) is different for the target unit from the alpha round to the beta round. In fact, by using the multiplication model 318, the accelerated genotype imputation system 106 determines a value of the column representing the target marker variant by performing a multiplication operation 334a for the alpha round, and determines another value of the column representing the target marker variant by performing a multiplication operation 334a for the beta round. In addition, by using the multiplication model 318, the accelerated genotype imputation system 106 determines a value for each row and each column by performing a multiplication operation 334b ​​for the alpha round, and determines another value for each row and each column by performing a multiplication operation 334b ​​for the beta round.

[0110] In contrast to the multiplication model 318, when the haploid model 320 is used, the accelerated genotype imputation system 106 performs a multiplication operation 334d as part of determining the intermediate allele probability 332b of the target unit. As an overview, the accelerated genotype imputation system 106 performs a multiplication operation 334d by multiplying the first transition-aware allele probability factor 338 with the adjacent marker intermediate allele probability 328b. By further performing a summation operation 340b to sum the adjacent marker transition factor-aware allele probability 342 and the summed adjacent marker transition-aware allele probability factor 336, the accelerated genotype imputation system 106 determines the intermediate allele probability 332b of the target unit.

[0111] In some embodiments, by using Figure 3B In the single-multiplication model 320 shown, the accelerated genotype inference system 106 selects a haplotype allele 330b for a column representing a target variant marker within a haplotype matrix. In some cases, the haplotype allele 330b takes the form of S bits, which select a value representing the haplotype allele to pass to downstream logic. For example, in certain embodiments, the accelerated genotype inference system 106 selects a haplotype allele 330b by identifying (i) an allele probability factor (e.g., Q0) corresponding to a sample reference haplotype allele of the target haplotype represented by a row or (ii) another allele probability factor (e.g., Q1) corresponding to a sample alternative haplotype of the target haplotype. Based on the identified allele probability factor (e.g., Q0 or Q1), the accelerated genotype inference system 106 transmits or sends a corresponding value representing the haplotype allele downstream for use in the summed adjacent marker transition-aware allele probability factor 336 and the first transition-aware allele probability factor 338. In fact, as Figure 3BAs further shown, the accelerated genotype imputation system 106 uses the selected haplotype allele 330 b as part of the summed neighboring marker transition-aware allele likelihood factor 336 and uses the first transition-aware allele likelihood factor 338 as part of the single-multiplicative model 320 .

[0112] As suggested above, in some embodiments, the accelerated genotype imputation system 106 predetermines a first transition-aware allele likelihood factor 338 and a second transition-aware allele likelihood factor (the latter as part of the summed neighboring marker transition-aware allele likelihood factors 336) before determining the median allele likelihood for a column representing a target marker variant within the haplotype matrix. To predetermine the first transition-aware allele likelihood factor 338, in some embodiments, the accelerated genotype imputation system 106 multiplies the allele likelihood factor (e.g., Q[m][Allele]) for a particular type of haplotype allele corresponding to haplotype allele 330b by a transition constant coefficient (P0) for transitioning between haplotypes from a haplotype reference set. To predetermine the summed neighboring marker transition-aware allele likelihood factor 336, the accelerated genotype inference system 106 multiplies the allele likelihood factor (e.g., Q[m][Allele]), the conversion linear coefficient for transitioning between haplotypes from a haplotype reference group (e.g., P1), and the summed neighboring marker intermediate allele likelihood 324 (e.g., Sum'[m-1]) of the neighboring marker variants.

[0113] During the rounds of the haplotype matrix, the accelerated genotype imputation system 106 also determines the neighboring marker intermediate allele probabilities 328b for neighboring cells representing variant markers and target haplotypes. In fact, in some embodiments, when the accelerated genotype imputation system 106 performs rounds of determining the intermediate allele probabilities of the haplotype matrix column by column, the accelerated genotype imputation system 106 determines the neighboring marker intermediate allele probabilities 328b for neighboring cells before reaching the target cell.

[0114] After predetermining the first transition-aware allele likelihood factor 338 and the neighboring marker intermediate allele likelihood 328b, the accelerated genotype estimation system 106 can perform a single round of parallel multiplication operations on the target unit. Figure 3BAs shown, the accelerated genotype imputation system 106 performs a multiplication operation 334d by multiplying the first transition-aware allele likelihood factor 338 (e.g., Q[m][Allele]*P1[m]) by the adjacent marker intermediate allele likelihood 328b (e.g., A'[m-1][k]). As an output of the multiplication operation 334d, the accelerated genotype imputation system 106 generates an adjacent marker transition factor-aware allele likelihood 342 (e.g., Q[m][Allele]*P1[m]*A'[m-1]).

[0115] like Figure 3B As further shown, the accelerated genotype imputation system 106 further determines the intermediate allele probability 332b of the target unit by performing a summation operation 340b. Specifically, the accelerated genotype imputation system sums the neighboring marker transition factor-aware allele probability 342 (e.g., Q[m][Allele]*P1[m]*A'[m-1]) and the summed neighboring marker transition-aware allele probability factor 336 (e.g., Q[m][Allele]*P0[m]*Sum'[m-1]) to determine the intermediate allele probability 332b (e.g., A[m][k]).

[0116] As suggested above, by performing three multiplication operations 334a-334c for each target cell using the continuous multiplication model 318, the accelerated genotype imputation system 106 will perform 3,000 multiplication operations for each row representing a haplotype from the haplotype reference group. In contrast, by performing multiplication operation 334d for each target cell using the single multiplication model 320, the accelerated genotype imputation system 106 reduces processing to approximately 1,000 multiplication operations for each row representing a haplotype from the haplotype reference group. Since multiplication operations on a configurable processor (such as an FPGA) consume considerable processing, the single multiplication model 320 significantly reduces both time and computer processing to determine the intermediate allele likelihood and output the allele likelihood.

[0117] In some embodiments, in addition to or as an alternative to performing a single round of parallel multiplication operations on target units, the accelerated genotype imputation system 106 can store and use the intermediate allele probability subset to hot-start determining certain intermediate allele probabilities during rounds across the haplotype matrix. According to one or more embodiments, Figure 4A Depicted is the accelerated genotype imputation system 106 storing and accessing a subset of intermediate allele likelihoods corresponding to a set of marker variants to hot start intermediate allele likelihood determinations during one or more rounds across a haplotype matrix. Figure 4BDepicted is an accelerated genotype imputation system 106 that (i) determines and stores a subset of intermediate allele probabilities corresponding to columns of marker variants that are grouped together, and (ii) generates a set of intermediate allele probabilities for rounds across a haplotype matrix by using the subset of intermediate allele probabilities as hot start points.

[0118] like Figure 4A As shown in , in some embodiments, the accelerated genotype imputation system 106 uses a configurable processor 400 to perform a sacrificial first round 402 of determining intermediate allele probabilities across cells of a haplotype matrix 404. The present disclosure refers to the sacrificial first round 402 as "sacrificial" because the accelerated genotype imputation system 106 performs the sacrificial first round 402 for the purpose of determining a subset of first round intermediate allele probabilities 406 corresponding to a subset of marker variants. In addition to the hot start point for regenerating the first round intermediate allele probabilities, in some embodiments, the accelerated genotype imputation system 106 does not directly use the intermediate allele probabilities determined during the sacrificial first round 402.

[0119] When performing the sacrifice of the first round 402, the accelerated genotype imputation system 106 may perform a forward round or a reverse round (or an alpha round or a beta round). As suggested above, in the forward round, the accelerated genotype imputation system 106 generates a forward intermediate allele possibility for a genomic region including a haplotype allele. On the contrary, in the reverse round, the accelerated genotype imputation system 106 generates a reverse intermediate allele possibility for a genomic region including a haplotype allele. Since the accelerated genotype imputation system 106 performs both a forward round (e.g., the second round) and a reverse round (e.g., the first round) as the basis for generating allele possibilities, regardless of the direction of the sacrifice round, the direction of the sacrifice round should not affect the allele possibilities (e.g., R0, R1). Regardless of the orientation, in some embodiments, the accelerated genotype imputation system 106 performs a sacrificial first round 402 by determining, cell by cell and column by column, an intermediate allele likelihood for each cell representing a combination of a marker variant and a haplotype from a haplotype reference set in the haplotype matrix 404. By performing the sacrificial first round 402, the accelerated genotype imputation system 106 utilizes the configurable processor 400 to determine, given a set of marker variants, a first round of intermediate allele likelihoods for genomic regions from a genomic sample that includes haplotype alleles corresponding to a set of haplotypes.

[0120] After executing the first round of sacrifice 402, if Figure 4AAs further shown, the accelerated genotype imputation system 106 identifies first-round intermediate allele probabilities 406a-406n from the first-round intermediate alleles determined according to the sacrificed first round 402. For example, in some embodiments, the accelerated genotype imputation system 106 identifies groups of marker variants (such as groups of 20, 100, 500, or 1,000 marker variants), and (ii) selects first-round intermediate allele probabilities from each group of marker variants to include in the subset of first-round intermediate allele probabilities 406. Thus, in some embodiments, the accelerated genotype imputation system 106 selects the intermediate allele probabilities of one column of marker variants for every 20, 100, 500, or 1,000 columns of marker variants within the haplotype matrix 404.

[0121] like Figure 4A As shown, first round intermediate allele likelihood 406a-406n represents the intermediate allele likelihood of the columns selected from each threshold number of columns representing the marker variant group. First round intermediate allele likelihood 406a, 406b and up to 406n together constitute a subset of first round intermediate allele likelihood 406.

[0122] In addition to identifying a subset of the first round intermediate allele possibilities 406, as Figure 4A As further shown, the accelerated genotype imputation system 106 stores a subset of the first round intermediate allele probabilities 406 on a memory device 408. As suggested above, the values ​​in the haplotype matrix 404 after sacrificing the first round 402 will be saturated or prove too many to be stored in the on-chip memory of the configurable processor 400. In order to reduce and redistribute the large amount of data of the haplotype matrix 404 after sacrificing the first round 402, the accelerated genotype imputation system 106 stores a subset of the first round intermediate allele probabilities 406 on a DRAM, SRAM, or other suitable memory for the storage device 408. The memory device 408 can be on-chip with the configurable processor 400 or off-chip with the configurable processor 400. Without saturating the memory of the configurable processor 400, the accelerated genotype imputation system 106 can access the subset of the first round intermediate allele probabilities 406 from the memory device 408 as a hot start point for determining the intermediate allele probabilities in the first round 410.

[0123] like Figure 4AAs further shown, in some embodiments, the accelerated genotype imputation system 106 regenerates the first-round intermediate allele probabilities from the sacrificed first round 402 by initializing allele probabilities determination at the marker variant set using a subset of the first-round intermediate allele probabilities 406. Specifically, when performing the first round 410, the accelerated genotype imputation system 106 (i) uses one of the first-round intermediate allele probabilities 406a-406n as the intermediate allele probabilities for a column of marker variants for every 20, 100, 500, or 1,000 columns of marker variants, and (ii) uses one of the first-round intermediate allele probabilities 406a-406n as a hot start point to determine subsequent intermediate allele probabilities in subsequent columns during the first round 410.

[0124] like Figure 4A As further shown, the accelerated genotype imputation system 106 may also perform a second round 412 to determine the second round of intermediate allele probabilities in a different direction than the first round 410. Specifically, the accelerated genotype imputation system 106 determines the second round of intermediate allele probabilities for genomic regions including haplotype alleles corresponding to the haplotype set given the set of marker variants using the configurable processor 400. Based on the regenerated first round intermediate allele probabilities and the second round intermediate allele probabilities, the accelerated genotype imputation system 106 generates allele probabilities for genomic regions including haplotype alleles.

[0125] Figure 4B A more detailed embodiment of an accelerated genotype imputation system 106 using an intermediate allele likelihood subset as a hot start point is depicted. Figure 4B As shown, the accelerated genotype imputation system 106 determines and stores a subset of the beta round intermediate allele probabilities 416 corresponding to the corresponding columns of the marker variants grouped by marker variant groups (including marker variant groups 1G-6G). The accelerated genotype imputation system 106 then accesses the subset of the beta round intermediate allele probabilities 416 and uses the separately stored intermediate allele probabilities as hot start points to generate intermediate allele probabilities in the alpha round and the beta round across the haplotype matrix 404.

[0126] like Figure 4B As shown, as a sacrificial β round, the accelerated genotype imputation system 106 performs a continuous β round 414, which determines the β round intermediate allele likelihood corresponding to the haplotype set and the marker variant set represented by the haplotype matrix 404. Specifically, the accelerated genotype imputation system 106 performs the continuous β round 414 by determining the β round intermediate allele likelihood for each cell within the haplotype matrix 404. Although Figure 4BThe continuous β rounds 414 are used as example sacrifice rounds, but the accelerated genotype estimation system 106 can also use continuous α rounds as sacrifice rounds. However, due to space constraints, Figure 4B Depicted are successive beta rounds 414 in horizontal blocks. The successive beta rounds 414 generate beta rounds of intermediate allele likelihoods (also referred to as beta values) for each cell and column of cells within the haplotype matrix 404. Although the successive beta rounds 414 are generally performed in opposite directions (and generally represented from right to left) across the haplotype matrix 404, Figure 4B Column groups 6G-1G are depicted representing groups of marker variants in reverse numerical order along a horizontal processing timeline.

[0127] After executing the continuous beta rounds 414, the accelerated genotype imputation system 106 identifies the beta round intermediate allele probabilities 416a-416e and stores them in the memory device 408 as a subset of the beta round intermediate allele probabilities 416. Figure 4B As indicated, each of the β round intermediate allele probabilities 416a-416e corresponds to a column representing a marker variant from a column group (e.g., one of the groups 1G-5G). For example, the β round intermediate allele probability 416a represents a column of intermediate allele probability values ​​selected from the column group 5G representing the marker variant group. In contrast, the β round intermediate allele probability 416b represents a column of intermediate allele probability values ​​selected from the column group 4G representing the marker variant group. The β round intermediate allele probabilities 416c, 416d, and 416e each similarly represent a column of intermediate allele probability values ​​selected from one of the column groups 3G, 2G, and 1G representing different marker variant groups. In some cases, the accelerated genotype imputation system 106 selects the last column of the intermediate allele probability (e.g., the β round intermediate allele probability 416e) as the β round intermediate allele probability to store a specific group (e.g., 1G) of column / marker variants.

[0128] like Figure 4BAs further shown, after storing the beta round intermediate allele probabilities 416a-416e as a subset of the beta round intermediate allele probabilities 416 in the memory device 408, the accelerated genotype imputation system 106 performs a segmented beta round 417. When performing the segmented beta round 417, the accelerated genotype imputation system 106 regenerates the intermediate allele probabilities values ​​determined in the consecutive beta rounds 414. However, in order to save memory on the chip for the configurable processor or other processor, the accelerated genotype imputation system 106 loads the beta round intermediate allele probabilities from the subset of the beta round intermediate allele probabilities 416 at certain columns to initialize (or warm-start) the determination of the beta round intermediate allele probabilities of the adjacent columns without re-determining the subset of the beta round intermediate allele probabilities 416 during the segmented beta round 417.

[0129] like Figure 4B As further depicted in FIG. 4 , the accelerated genotype imputation system 106 loads the associated stored subset of the beta round intermediate allele likelihoods in the associated columns during the segmented beta round 417. Although the segmented beta round 417 is typically performed across the haplotype matrix 404 in the opposite direction (and typically represented from right to left), Figure 4B Depicted are groups of columns representing groups of marker variants progressing in reverse numerical order along a horizontal processing timeline.As suggested above, if the accelerated genotype imputation system 106 were to perform a sacrificial alpha round instead of or in addition to a sacrificial beta round, the accelerated genotype imputation system 106 would also perform a segmented alpha round.

[0130] To illustrate the sequence of the segmented β round 417, in some embodiments, the accelerated genotype imputation system 106 determines the β round intermediate allele probability of the initial column group 0G, and then loads the β round intermediate allele probability 416e of the first column of the first column group 1G. Based on the β round intermediate allele probability 416e, the accelerated genotype imputation system 106 determines the β round intermediate allele probability of the column adjacent to the first column within the first group of columns 1G. Similarly, the accelerated genotype imputation system 106 determines the β round intermediate allele probability of the entire first column group 1G, and then loads the β round intermediate allele probability 416d of the first column of the second column group 2G. Based on the β round intermediate allele probability 416d, the accelerated genotype imputation system 106 determines the β round intermediate allele probability of the column adjacent to the first column within the second group of columns 2G.

[0131] In addition to the segmented beta round 417, such as Figure 4BAs further shown, the accelerated genotype imputation system 106 also performs a continuous alpha round 418, which determines the alpha round median allele likelihood corresponding to the set of haplotypes and the set of marker variants represented by the haplotype matrix 404. Specifically, the accelerated genotype imputation system 106 performs the continuous alpha round 418 by determining the alpha round median allele likelihood for each cell within the haplotype matrix 404. Since the continuous alpha round 418 is generally performed in a forward direction (and generally represented from left to right) across the haplotype matrix 404, Figure 4B Column groups 0G-6G are depicted representing groups of marker variants in numerical order along a horizontal processing timeline.

[0132] like Figure 4B As further shown, as both the segmented beta rounds 417 and the continuous alpha rounds 418 progress, the accelerated genotype imputation system 106 determines the segmented allele probabilities 420. To illustrate the sequence of the segmented allele probabilities 420, in some embodiments, the accelerated genotype imputation system 106 determines the allele probabilities for the initial column group 0G by multiplying the sum of the corresponding beta round and alpha round intermediate allele probabilities. When the accelerated genotype imputation system 106 subsequently loads the beta round intermediate allele probabilities 416e of the first column in the first column group 1G and determines the alpha round intermediate allele probabilities for the first column as part of the continuous alpha round 418, the accelerated genotype imputation system 106 multiplies the corresponding sums of the beta round intermediate allele probabilities 416e and the alpha round intermediate allele probabilities for the first column in the first column group 1G. Based on this multiplication of the sums, the accelerated genotype imputation system 106 determines the allele probabilities (R0 and R1) for the first column in the first column group 1G. In some embodiments, the accelerated genotype imputation system 106 overwrites the corresponding sum of the beta round intermediate allele likelihood 416e and the alpha round intermediate allele likelihood of the first column in the first column group 1G with the allele likelihood of the first column in the first column group 1G.

[0133] As further illustration, in some embodiments, the accelerated genotype imputation system 106 determines the allele probabilities of the first column group 1G by multiplying the sum of the corresponding beta round intermediate allele probabilities and the alpha round intermediate allele probabilities. When the accelerated genotype imputation system 106 loads the beta round intermediate allele probabilities 416d of the first column in the second column group 2G and determines the alpha round intermediate allele probabilities of the first column as part of the continuous alpha round 418, the accelerated genotype imputation system 106 multiplies the corresponding sums of the beta round intermediate allele probabilities 416d and the alpha round intermediate allele probabilities of the first column in the second column group 2G. Based on this multiplication of the sums, the accelerated genotype imputation system 106 determines the allele probabilities (R0 and R1) of the first column in the second column group 2G, and (in some cases) overwrites the corresponding sums with the allele probabilities of the first column in the second column group 2G.

[0134] In addition to or in lieu of using the intermediate allele likelihood subset as a hot start point, in some embodiments, the accelerated genotype imputation system 106 determines and uses a running sum of the intermediate allele likelihoods to speed up the rounds of determining the intermediate allele likelihoods across the haplotype matrix. According to one or more embodiments, Figure 5A An accelerated genotype imputation system 106 is depicted that determines a running sum of intermediate allele likelihoods for genomic regions exhibiting haplotype alleles for one or more haplotypes in column n-1 (representing a first marker variant) as a running input for determining individual intermediate allele likelihoods for genomic regions exhibiting haplotype alleles in column n (representing a second marker variant). Figure 5B Depicted is a comparison of the accelerated genotype imputation system 106 using a full sum model and a running sum model to determine the column sum of intermediate probabilities and the impact of such models on waiting time periods.

[0135] like Figure 5AAs shown, the accelerated genotype inference system 106 executes the full column sum model 502 to determine the intermediate allele possibility of the column representing different variant markers. When the full column sum model 502 is executed, for example, the accelerated genotype inference system 106 determines the sum of the intermediate allele possibility 506 of the column n representing the second marker variant before determining the intermediate allele possibility 508 of the column n+1 representing the third marker variant. When the full column sum model 502 is executed, the full column sum model 502 causes the processor to wait for the waiting time cycle to determine the intermediate allele possibility 506 of the column n- and generate the allele possibility of the column n before starting to determine the intermediate allele possibility 508 of the column n+1. Since the haplotype matrix may need to determine the value equivalent to millions, billions or tens of thousands of cells, and the parallel determination of the intermediate allele possibility of the cell is more effective than the serial method, it is proved that such waiting time cycle costs are high, and significantly slows down the process of phasing and inferring the haplotype allele possibility of the single marker allele corresponding to the genomic region by an average of about 17.5 hours.

[0136] In contrast to the full column sum model 502, in some embodiments, the accelerated genotype imputation system 106 performs a run column sum model 504 to determine the median allele likelihood for columns representing different variant markers. Figure 5A As shown, for example, the accelerated genotype imputation system 106 determines a running sum of the intermediate allele likelihoods 510 for genomic regions that exhibit haplotype alleles given one or more haplotypes of a first marker variant represented by column n-1. By using the running sum of the intermediate allele likelihoods 510 as a running input for column n, the accelerated genotype imputation system 106 determines a sum of the intermediate allele likelihoods 512 for genomic regions that exhibit haplotype alleles given a second marker variant represented by column n.

[0137] When executing the run column sum model 504, the accelerated genotype estimation system 106 speeds up the parallel determination of the intermediate allele likelihoods for the haplotype matrix cells. Figure 5AAs further shown, the accelerated genotype imputation system 106 further determines such a running sum of the intermediate allele likelihoods given the second marker variant represented by column n. By using the running sum of the intermediate allele likelihoods of column n as a running input, the accelerated genotype imputation system 106 similarly determines the intermediate allele likelihoods 514 of the genomic region exhibiting the haplotype alleles given the third marker variant represented by column n+1. In fact, the accelerated genotype imputation system 106 can derive (or otherwise determine) the intermediate allele likelihoods 512 of column n+1 based on the sum of the intermediate allele likelihoods 514 of column n. Unlike the full column sum model 502, by using the running column sum model 504, the accelerated genotype imputation system 106 does not need to wait to determine the sum of the intermediate allele likelihoods of one column before determining the individual allele likelihoods (or sums) of another column within the haplotype matrix.

[0138] Figure 5B A comparison of the accelerated genotype imputation system 106 executing the full column sum model 502 and running the column sum model 504 is depicted in greater detail with respect to the relative timing of the input values ​​and output values ​​for each column. When executing the full column sum model 502, the accelerated genotype imputation system 106 may perform Figure 3B 334a, 334b, and 334c and summation operation 340a as depicted in and described above determine the median allele likelihood (e.g., A[m][k]) for the target unit. In fact, the disclosure refers to the full column sum model 502 as a "full column sum" because one such multiplication operation requires summing the entire column of the median allele likelihood (e.g., A[m][k] values). Specifically, when performing Figure 3B In the multiplication operation 334a depicted in FIG. 1 , in some embodiments, the accelerated genotype imputation system 106 multiplies the transition constant coefficient (P0) of the column representing the target marker variant by the normalized summed neighboring marker intermediate allele likelihood (Sum'[m-1]) of the neighboring marker variants represented by the column. Since the summed neighboring marker intermediate allele likelihood (Sum'[m-1]) requires summing the intermediate allele likelihoods of the entire column representing the neighboring marker variants within the haplotype matrix, the full column sum model 502 is applied on the processor. Figure 5B The wait time period depicted in is used to determine and sum the median allele likelihood for the columns representing target marker variants without performing other parallel operations.

[0139] like Figure 5BAs shown, when executing the full-column sum model 502, the accelerated genotype imputation system 106 inputs the per-unit column input value 516a into the cell of column n-1 to determine the per-unit column output value 518a of column n-1. As suggested above, in some embodiments of the full-column sum model 502, the per-unit column input value 516a includes the allele probability factor (Q0 or Q1) for each cell in column n-1, the conversion coefficient (P1[m] and P0[m]), the summed adjacent marker intermediate allele probability (Sum'[m-1]) and the normalization value (Norm[m-1]). Based on the per-unit column input value 516a, in some embodiments, the accelerated genotype imputation system 106 determines the per-unit column output value 518a in the form of the intermediate allele probability, which is represented as the α value (e.g., A[m][k] value) or β value (e.g., B[m][k] value) for the α round or the β round, respectively. Figure 5B The time to determine the per-cell column output value 518a from the cell of column n-1 is depicted as the cell update latency period 524.

[0140] Based on the per-cell column output value 518a, the accelerated genotype imputation system 106 determines a column sum output value 520a for column n-1 as part of the full column sum model 502. For example, in some embodiments, the accelerated genotype imputation system 106 determines the alpha value for column n-1 The sum of the values ​​of column n-1 and the beta value Based on the column sum output value 520a, the accelerated genotype estimation system 106 determines the probability 522a of each column allele for column n-1. For example, the accelerated genotype estimation system 106 calculates the α value of column n-1. The sum and beta value of The sum of is multiplied to determine the allele likelihood (R0 and R1) for column n-1.

[0141] When executing the full column sum model 502, such as Figure 5B As shown, the accelerated genotype imputation system 106 determines the column sum output value 520a and the per-column allele likelihood 522a before inputting each column input value 516b into the cell of column n to determine the per-cell column output value 518b of column n. Since the processor of the accelerated genotype imputation system 106 determines the column sum output value 520a and the per-column allele likelihood 522a before inputting each column input value 516b, the full column sum model 502 is created. Figure 5B526a and each column allele likelihood wait time period 528a are depicted in FIG. In other words, the full column sum model 502 requires the processor to wait through a certain wait time period to perform the following two operations (summing the intermediate allele likelihoods of adjacent markers and generating allele likelihoods) without performing other parallel operations on adjacent haplotype matrix columns.

[0142] like Figure 5B As further shown, the full column sum model 502 similarly creates a column sum wait time between column n and column n+1 and a per-column allele likelihood wait time. The accelerated genotype imputation system 106 determines the column sum output value 520b and the per-column allele likelihood 522b before inputting the per-column input value 516c into the cell of column n+1 to determine the per-cell column output value 518c of column n+1. Since the processor of the accelerated genotype imputation system 106 determines the column sum output value 520b and the per-column allele likelihood 522b before inputting the per-column input value 516c, just like other haplotype matrix columns, the full column sum model 502 also creates the column sum wait time 526b and the per-column allele likelihood wait time 528b.

[0143] In contrast to the full column sum model 502, the accelerated genotype imputation system 106 eliminates such empty waiting time periods when executing the run column sum model 504. For example, in some embodiments, the accelerated genotype imputation system 106 determines, for column n-1 representing neighboring marker variants, a median allele likelihood (e.g., ∑ k(S＝＝0) Similarly, the accelerated genotype imputation system 106 determines, for column n-1 representing neighboring marker variants, a median allele likelihood (e.g., ∑ k(s＝＝1) As indicated above, in some cases, the first type of haplotype allele comprises a sample reference haplotype allele (e.g., S[k][m] value is 0), and the second type of haplotype allele comprises a sample alternative haplotype allele (e.g., S[k][m] value is 1).

[0144] Based on the running sum of the first subset of intermediate allele probabilities and the running sum of the second subset of intermediate allele probabilities, for column n representing the target marker variant, the accelerated genotype imputation system 106 determines the sum of the intermediate allele probabilities for the genomic region including the haplotype alleles of the haplotypes from the haplotype reference group (e.g., Sum[m]). For example, in some embodiments, the accelerated genotype imputation system 106 determines the sum of the intermediate allele probabilities from the alpha round and the sum of the intermediate allele probabilities from the beta round. Based on the sum of the intermediate allele probabilities, for column n representing the target marker variant, the accelerated genotype imputation system 106 generates the allele probabilities (R0 and R1) for the genomic region including the haplotype alleles.

[0145] As described above, the accelerated genotype imputation system 106 may predetermine certain variables before a round of the haplotype matrix to speed up the round. In some cases, for example, the accelerated genotype imputation system 106 predetermines and considers various per-unit column input values ​​as part of the running column sum model 504. For example, in some embodiments, the accelerated genotype imputation system 106 predetermines a first transition-aware allele likelihood factor (e.g., Q0[m]*P0[m]*(K-S1)) corresponding to a row of haplotype alleles of the first type and a second transition-aware allele likelihood factor (e.g., Q1[m]*P0[m]*S1) corresponding to a row of haplotype alleles of the second type. Therefore, in addition to the running sum, the accelerated genotype imputation system 106 may further determine the sum of the intermediate allele likelihoods (e.g., Sum[m]) based on the first transition-aware allele likelihood factor corresponding to the row of haplotype alleles of the first type and the second transition-aware allele likelihood factor corresponding to the row of haplotype alleles of the second type.

[0146] As another example, and as indicated above, the accelerated genotype imputation system 106 can estimate the neighboring marker sum of the intermediate allele likelihood (e.g., Sum[m-1]) instead of summing all neighboring marker intermediate allele possibilities (e.g., A[m]1][k] values) to determine the neighboring marker sum of the intermediate allele likelihood (e.g., Sum[m-1]). Thus, in some embodiments, for column n-1 representing neighboring marker variants, the accelerated genotype imputation system 106 estimates the running sum of the first subset of the intermediate allele possibilities (e.g., ∑ k(S＝＝0) A[m-1][k]) and a running sum of a second subset of intermediate allele possibilities (e.g., ∑ k(s＝＝1) A[m-1][k]) determines the sum of neighboring markers (e.g., Sum[m-1]) that determine the median allele likelihood for the genomic region that includes the haplotype allele.

[0147] Thus, in some embodiments, for column n representing marker variants, the accelerated genotype imputation system 106 determines a sum of intermediate allele probabilities (Sum[m]) based on a combination of: (i) the neighboring marker sum of the intermediate allele probabilities, (ii) a first transition-aware allele probability factor for rows corresponding to a first type of haplotype allele, (iii) a running sum of a first subset of the intermediate allele probabilities, (iv) a running sum of a second subset of the intermediate allele probabilities, and (v) a second transition-aware allele probability factor for rows corresponding to a second type of haplotype allele. In some such cases, for example, the accelerated genotype inference system 106 determines the product of the neighboring marker sum of the intermediate allele likelihood (Sum[m-1]) and the first transition-aware allele likelihood factor (Q0[m]*P0[m]*(K-S1)) for the row corresponding to the first type of haplotype allele, and adds the product to the second transition-aware allele likelihood factor (Q1[m]*P0[m]*S1) for the row corresponding to the second type of haplotype allele.

[0148] In addition to determining the running sum, in some cases, the accelerated genotype imputation system 106 also multiplies the running sum of a subset of the intermediate allele probabilities by the transition-aware allele probability factor as part of determining the sum (Sum[m]) of the intermediate allele probabilities for column n representing the target marker variants. For example, in some embodiments, the accelerated genotype imputation system 106(i) multiplies the running sum of the first subset of the intermediate allele probabilities by the first transition-aware allele probability factor (e.g., Q0[m]*P1[m]*∑ k(s＝＝0) A[m-1][k]) and multiplying the running sum of the second subset of intermediate allele probabilities by a second transition-aware allele probability factor (e.g., Q1[m]*P1[m]*∑ k(s＝＝1) Based on the running sum of the multiplication of the first subset of intermediate allele probabilities and the running sum of the multiplication of the second subset of intermediate allele probabilities, the accelerated genotype imputation system 106 determines the sum of the intermediate allele probabilities (Sum[m]) for column n representing target marker variants.

[0149] Therefore, in some embodiments, for column n representing marker variants, the accelerated genotype imputation system 106 determines the sum of intermediate allele probabilities (Sum[m]) by summing the products of (a) a running sum of the multiplication of a first subset of intermediate allele probabilities, (b) a running sum of the multiplication of a second subset of intermediate allele probabilities, and (c) (i) the normalized values ​​of neighboring marker variants, (ii) the product of the neighboring marker sums of intermediate allele probabilities, and (iii) the sum of a first transition-aware allele likelihood factor for rows corresponding to a first type of haplotype allele and a second transition-aware allele likelihood factor for rows corresponding to a second type of haplotype allele.

[0150] Figure 5B The effect of running the column sum model 504 on various waiting time periods is illustrated. Figure 5B As shown, when executing the run column sum model 504, the accelerated genotype imputation system 106 inputs the per-cell column input value 530b into the cells of column n to determine the per-cell column output value 532b from the cells of column n. As suggested above, in some embodiments of running the column sum model 504, the per-cell column input value 530b for each cell in column n includes: (i) a normalized value of neighboring marker variants (e.g., Norm[m-1]); (ii) an estimated neighboring marker sum of intermediate allele likelihoods (e.g., Sum[m-1]); (iii) a first transition-aware allele likelihood factor for rows corresponding to a first type of haplotype allele (e.g., Q0[m]*P0[m]*(K-S1)); (iv) a second transition-aware allele likelihood factor for rows corresponding to a second type of haplotype allele (e.g., Q1[m]*P0[m]*S1); (v) a multiplicative sum of a first subset of intermediate allele likelihoods (e.g., Q0[m]*{1[m]*∑ k(S＝＝0) A[m-1][k]); and (v) the multiplicative sum of the second subset of intermediate allele possibilities (e.g., Q1[m]*P1[m]*∑ k(S＝＝1) Based on the per-unit column input value 530b, in some embodiments, the accelerated genotype imputation system 106 determines a per-unit column output value 532b in the form of an intermediate allele likelihood, which is represented as an α value (e.g., A[m-1][k] value) or a β value (e.g., B[m][k] value) for the α round or the β round, respectively.

[0151] Since the column sum model 504 is run, as Figure 5BAs shown, the accelerated genotype estimation system 106 determines the column sum output value 534b (e.g., Sum[m]) for column n as the sum of the intermediate allele probabilities before completing the determination of each intermediate allele probability in the per-unit column output value 532b. In fact, Figure 5B As further indicated, the accelerated genotype imputation system 106 determines the column sum output value 534b while also determining the column sum output value 534a for column n-1 and the per-column allele likelihood 536a for column n-1. Thus, the accelerated genotype imputation system 106 inputs the per-unit column input value 530b and determines the column sum output value 534b during (i) the column sum wait time 540 for column n-1 and (ii) the per-column allele likelihood wait time 542 for column n-1. Thus, running the column sum model 504 ensures that the processor of the accelerated genotype imputation system 106 determines the intermediate allele likelihood for column n during the column sum wait time 540 for column n-1 and the per-column allele likelihood wait time 542 for column n-1 (rather than waiting through the wait time).

[0152] like Figure 5B As further indicated, in some embodiments, the accelerated genotype imputation system 106 applies the run column sum model 504 to column n-1 and column n+1. For example, the accelerated genotype imputation system 106 inputs the per-unit column input value 530c of column n+1 and determines the column sum output value 532c of column n+1, while also determining the column sum output value 534b of column n and the per-column allele likelihood 536b of column n, thereby ensuring that the processor of the accelerated genotype imputation system 106 does not wait through the column sum wait time and the per-column allele likelihood wait time of column n without performing parallel operations on other columns.

[0153] Although not in Figure 5B , but in some embodiments, the accelerated genotype imputation system 106 can use the running sum of the intermediate allele probabilities from column n-2 to determine the sum of the intermediate allele probabilities for column n-1. Specifically, the accelerated genotype imputation system 106 inputs the per-unit column input value 530a for column n-1 and determines the column sum output value 534a for column n-1, while also determining the column sum output value for column n-2 and the per-column allele probabilities for column n-2. Therefore, although Figure 5B A cell update wait time 538 representing the time to determine the per-cell column output value 532a for the cell from column n−1 and the processing is depicted, but in some cases, the accelerated genotype imputation system 106 uses a processor of the accelerated genotype imputation system 106 to determine other values ​​of the haplotype matrix during the cell update wait time 538 .

[0154] As described above, in some embodiments, the accelerated genotype imputation system 106 intelligently transmits data to increase throughput on a configurable processor or other processor. According to one or more embodiments, Figure 6 The accelerated genotype imputation system 106 is illustrated as storing haplotype allele indicator numbers for a haplotype matrix on a memory device and accessing the stored haplotype allele indicator data to determine values ​​as part of a pass across the haplotype matrix.

[0155] As described above, HMM-based genotype inference may need to determine and store a large amount of data, such as the values ​​of millions, billions or trillions of cells in a haplotype matrix. For example, in some embodiments, the accelerated genotype inference system 106 inputs the value representing the haplotype allele into each cell of the haplotype matrix, such as (i) one "S" bit indicates a sample reference haplotype allele of a specific haplotype, and (ii) another "S" bit indicates a sample replacement haplotype allele of a specific haplotype. As described above, the present disclosure refers to such input values ​​representing haplotype alleles as haplotype allele indicator data of a haplotype matrix. Since the haplotype allele indicator data of a haplotype matrix with millions, billions or trillions of cells may consume more than several gigabytes of memory, the haplotype allele indicator data occupies the bandwidth of a high-speed bus of a configurable processor (such as a fast peripheral component interconnect (PCIe)) or other interfaces connecting a processor card to other hardware in a computing device.

[0156] To save bandwidth on PCIe or other interfaces, such as Figure 6 As shown, the accelerated genotype imputation system 106 stores the haplotype allele indicator data 602a of the haplotype matrix on the memory device 600. In some cases, the accelerated genotype imputation system 106 stores the haplotype allele indicator data 602a on an on-chip DRAM, SRAM or other suitable memory. Since the haplotype allele indicator data 602a is easily accessible, the accelerated genotype imputation system 106 can access the haplotype allele indicator data 602a and transfer it from the memory device 600 to the configurable processor 604 to perform the transfer of the intermediate allele possibility across the haplotype matrix. For example, in some embodiments, the accelerated genotype imputation system 106 uses the configurable processor 604 to access the haplotype allele indicator data 602a of the haplotype matrix from the memory device 600 to generate the allele possibility of the genotype imputation model.

[0157] Since the haplotype allele indicator data 602a (or "S" bit data) is in the same format as the haploid or diploid genotype inference, the accelerated genotype inference system 106 can store and access the haplotype allele indicator data 602a for the hidden Markov haploid or diploid genotype inference model. Therefore, the accelerated genotype inference system 106 can use the configurable processor 604 to access the haplotype allele indicator data 602a of the haplotype matrix from the memory device 600 to generate allele possibilities using the hidden Markov haploid genotype inference model or the hidden Markov diploid genotype inference model. When input to the haplotype matrix of the round, Figure 6 The input data is depicted as haplotype allele indicator data 602 b on a matrix that is analyzed by a configurable processor 604 .

[0158] To perform approximately 40,000 HMM calculation tasks in approximately 60 seconds for a single processor thread, in some embodiments, the accelerated genotype imputation system 106 requires approximately 10 gigabytes per second of PCIe throughput, with a margin of 6 gigabytes per second available during a round. In some embodiments, the accelerated genotype imputation system 106 saves 4 gigabytes per second or more of PCIe bandwidth by storing the haplotype allele indicator data 602a of the haplotype matrix on and accessing from the on-chip DRAM (or other on-chip memory).

[0159] As described above, in some embodiments, the accelerated genotype imputation system 106 includes and uses a custom architecture to run a genotype determination model, such as GLIMPSE. According to one or more embodiments, Figure 7 An accelerated computing engine 700 is illustrated that includes various custom engines and memory devices to determine allele likelihood 722 using a genotype inference model. The following paragraphs describe various memory devices and operations for determining allele likelihood 722. Figure 7 The accelerated computing engine 700 depicted in FIG. 7 represents a memory device and engine for haploid HMM computations, but a similar accelerated computing engine may be used for diploid HMM computations.

[0160] like Figure 7As shown, for example, the accelerated computing engine 700 includes an alpha column memory 704a and a beta column memory 704b. In some embodiments, the alpha column memory 704a and the beta column memory 704b store pre-normalized intermediate allele probabilities for the alpha round and the beta round, respectively. Specifically, in certain embodiments, the alpha column memory 704a and the beta column memory 704b store a column of pre-normalized alpha values ​​(e.g., A[m][k] values) and a column of pre-normalized beta values ​​(e.g., B[m][k] values), respectively. In terms of format, the alpha column memory 704a and the beta column memory 704b may each store a K×Z ABwide The values ​​are organized in bits, that is, the K rows represent Z-bit wide haplotypes of the stored pre-normalized α or β values.

[0161] like Figure 7 As further shown, the accelerated computing engine 700 includes a haplotype allele indicator memory 708. The haplotype allele indicator memory 708 stores haplotype allele indicator data (or "S" bit data), which includes an input value representing the haplotype allele of each cell of the haplotype matrix. Figure 2B and Figure 6 In terms of format, the haplotype allele indicator memory 708 can store values ​​or bits of haplotype allele indicator data organized as M×K bits, that is, M columns representing marker variants and K rows representing haplotypes from a haplotype reference set. As described above, in some embodiments, the accelerated genotype imputation system 106 transfers the haplotype allele indicator data from an on-chip DRAM or another memory device to the haplotype allele indicator memory 708 to perform a round of the haplotype matrix.

[0162] In addition to the haplotype allele indicator memory 708, the accelerated computing engine 700 also includes a transformation coefficient memory 710. The transformation coefficient memory 710 stores transformation coefficients (e.g., P0 and P1 values) corresponding to columns or cells of the haplotype matrix. The transformation coefficient memory 710 can store the transformation coefficients organized as 2×M×Z p The values ​​of the transition coefficients of the bits, that is, two sections or blocks of values ​​in the M columns (e.g., one section for P0 values ​​and one section for P1 values), represent the Z of the input P0 values ​​and P1 values. p Flag variants in bit width.

[0163] In addition to the conversion coefficient memory 710, the accelerated computing engine 700 also includes an allele likelihood factor memory 712. The allele likelihood factor memory 712 stores the allele likelihood factors (e.g., Q0 and Q1 values) corresponding to the columns or cells of the haplotype matrix. The allele likelihood factor memory 712 can store the allele likelihood factors organized as 2×M×Z Q The values ​​of the allele likelihood factors of bits, that is, two bins or blocks of values ​​in the M columns (e.g., one bin for Q0 values ​​and one bin for Q1 values), representing the Z of the input Q0 values ​​and Q1 values Q Flag variants in bit width.

[0164] like Figure 7 As further shown, the accelerated computing engine 700 also includes an intermediate allele probability memory 716. The intermediate allele probability memory 716 stores the intermediate allele probability of the haplotype matrix. For example, in some cases, the intermediate allele probability memory 716 stores the α value and the β value determined across the full haplotype matrix. In terms of organization, the intermediate allele probability memory 716 can store the intermediate allele probability organized as W×K×Z AB The intermediate allele likelihood of bits, that is, W columns in the marker variant group, K rows representing haplotypes, and Z bits width for storing normalized alpha values ​​or beta values. Therefore, in some embodiments, the intermediate allele likelihood memory 716 organizes the alpha values ​​or beta values ​​by marker variant group to be compatible with a subset of the intermediate allele likelihoods of the rounds that initialize the intermediate allele likelihood at the hot start point.

[0165] In some embodiments, the accelerated genotype imputation system 106 determines the allele likelihood 722 for one or more of the cells, columns, or haplotype matrices using a custom architecture of the accelerated computing engine 700. Figure 7 As shown, for example, the accelerated computing engine 700 uses SNIFF 702a to generate an alpha normalized value, and uses SNIFF 702a to generate a beta normalized value. The accelerated computing engine 700 also applies the normalized value from SNIFF 702a to normalize 706 the neighboring marker intermediate likelihood values ​​from the alpha value column stored in the alpha column memory 704a. Similarly, the accelerated computing engine 700 applies the normalized value from SNIFF 702b to normalize 706b the neighboring marker intermediate likelihood values ​​from the beta value column stored in the beta column memory 704b.

[0166] like Figure 7As further shown, the accelerated computing engine 700 uses the union engine 714 to determine the intermediate probability value of the target unit with the haplotype matrix. Specifically, the accelerated computing engine 700 (i) indirectly receives the normalized neighboring marker intermediate allele probability from the α column memory 704a and the β column memory 704b, and (ii) combines the haplotype allele indicator from the haplotype allele indicator memory 708, the transformation coefficient from the transformation coefficient memory 710, and the allele probability factor from the allele probability factor memory 712 with the normalized neighboring marker intermediate allele probability to determine (iii) the intermediate allele probability of the target unit stored in the intermediate allele probability memory 716. The accelerated computing engine 700 also uses the allele probability engine 718 to determine the allele probability 722 of the target unit based on the intermediate allele probability stored in the intermediate allele probability memory 716.

[0167] like Figure 7 As further indicated in, in some embodiments, the accelerated computing engine 700 receives the intermediate allele possibility subset 720a corresponding to the marker variant group from the memory device. Consistent with the above disclosure, in some cases, the accelerated computing engine 700 uses the intermediate allele possibility subset to initialize the allele possibility determination at the marker variant group, thereby regenerating the first round of intermediate allele possibility. As further indicated, the accelerated computing engine 700 also performs the sacrifice of the first round and determines the intermediate allele possibility subset 720b corresponding to the marker variant group, which can be stored on the memory device and accessed later to initialize the allele possibility determination at the corresponding marker variant group.

[0168] In fact, in some cases, the accelerated genotype imputation system 106 can use the accelerated computation engine 700 to determine and access the intermediate allele likelihood subsets, as described above with respect to FIG. 4A to FIG. 4B In addition, in some embodiments, the accelerated genotype estimation system 106 uses the accelerated computing engine 700 to determine a single round of concurrent multiplication operations, determine and use a running sum of a subset of intermediate allele possibilities, or perform the above-mentioned FIG. 3A to FIG. 3B and FIG. 5A to FIG. 5B Other embodiments described.

[0169] In addition to the accelerated computing engine as part of the custom architecture, in some embodiments, the accelerated genotype estimation system 106 also includes a data flow engine that can assign HMM calculation tasks to the accelerated computing engines in the accelerated computing engine cluster and manage data communications with the central processing unit (CPU), memory, and accelerated computing engines. According to one or more embodiments, Figure 8A configurable processor board 800 is depicted, which includes a data flow engine 802, an accelerated computing engine cluster 804, and an on-board memory device 822 for executing a genotype imputation model. Figure 8 As depicted in , the data flow engine 802 interacts and interfaces with the accelerated computing engine cluster 804, the onboard memory device 822, and the CPU to control, distribute, or otherwise manage data for the HMM computing task. Although the following paragraphs describe the interaction and data exchange between the data flow engine 802 and the accelerated computing engine 804a in the accelerated computing engine cluster 804, the same interaction and data exchange can be performed by each of the data flow engine 802 and the accelerated computing engines 804b-804n.

[0170] like Figure 8 As indicated, for example, the configurable processor board 800 is a local server device (e.g., Figure 1 A portion of a local device 110 shown) or a sequencing device (e.g., Figure 1 As part of such a computing device, in some embodiments, the data flow engine 802 on the configurable processor board 800 includes a PCIe interface for the FPGA and a double data rate (DDR) interface to an onboard memory device 822 (such as a DRAM).

[0171] In addition to being used as an interface or as part of being used as an interface, in some embodiments, the data flow engine 802 sends data to and receives data from other hardware on the CPU, the onboard memory device 822, and the accelerated genotype imputation system 106 to determine the intermediate allele possibility, the allele possibility, or other HMM calculations. As part of CPU communication 818, in some embodiments, the data flow engine 802 receives data indicators from the CPU to perform genotype imputation for one or more genomic regions of the genomic sample based on the prior genotype possibility derived from the nucleotide fragment read. As part of memory communication 820, in some cases, the data flow engine 802 uses the onboard memory device 822 to send and receive input or output requests to store or access data for genotype imputation or phasing. Such requests may include, for example, sending or receiving a column of intermediate allele possibilities (e.g., a column of alpha values ​​or beta values) or an intermediate allele possibility subset as a hot start point.

[0172] As just indicated, in some embodiments, the accelerated genotype imputation system 106 can exchange the subset of intermediate allele probabilities as a warm start point between the data flow engine 802 and the onboard memory device 822. For example, in some cases, the accelerated genotype imputation system 106 (i) sends the subset of first-round intermediate allele probabilities from the onboard memory device 822 to the data flow engine 802, and (ii) sends the subset of first-round intermediate allele probabilities from the data flow engine 802 to the accelerated computing engine 804a in the accelerated computing engine cluster 804 to regenerate the first-round intermediate allele probabilities based on the subset of first-round intermediate allele probabilities.

[0173] In addition to CPU communication 818 and memory communication 820, in some embodiments, the data flow engine 802 distributes the HMM calculation tasks to the individual accelerated computing engines of the accelerated computing engine cluster 804. For illustration, in some cases, the data flow engine assigns a single HMM calculation task to a single accelerated computing engine in the accelerated computing engine cluster 804 for a haplotype matrix of approximately 50 million cells, resulting in approximately 40,000 haplotype calls. Although other HMM calculation tasks may be larger or smaller than the aforementioned examples, in some embodiments, each of the individual HMM calculation tasks includes input values ​​and output values ​​of such a haplotype matrix.

[0174] like Figure 8 As indicated in, for example, the data flow engine 802 may send input values ​​806 of a target column or haplotype matrix for genotype inference to the accelerated computing engine 804a or receive output values ​​808 for a target column or haplotype matrix from the accelerated computing engine 804a, such as allele probabilities or intermediate allele probabilities subsets. The data flow engine 802 may also (i) receive the intermediate allele probabilities subset 810b from the sacrificed first round of the accelerated computing engine 804a as a hot start point, or (ii) send the intermediate allele probabilities subset 810a to the accelerated computing engine 804a as a hot start point to regenerate the column of intermediate allele probabilities initially determined in the sacrificed first round.

[0175] As an example of input values ​​806 and output values ​​808, in some embodiments, the accelerated genotype imputation system 106 sends a corresponding input value set including allele likelihood factors, transition coefficients, and haplotype allele values ​​from the data flow engine 802 to a corresponding accelerated computing engine of the accelerated computing engine cluster 804. Based on the corresponding input value set, the corresponding accelerated computing engine determines a corresponding intermediate allele likelihood set corresponding to a corresponding subset of marker variants and a subset of haplotypes.

[0176] To further illustrate, in certain specific implementations, the accelerated genotype imputation system 106 (i) sends a first input value set including allele likelihood factors, transition coefficients, and haplotype allele values ​​from the data flow engine 802 to the accelerated computing engine 804a, and (ii) sends a second input value set including allele likelihood factors, transition coefficients, and haplotype allele values ​​from the data flow engine 802 to the accelerated computing engine 804b. Based on the first input value set, the accelerated computing engine 804a determines an intermediate first allele likelihood set corresponding to a first subset of marker variants and a first subset of haplotypes. Similarly, based on the second input value set, the accelerated computing engine 804b determines a second intermediate allele likelihood set corresponding to a second subset of marker variants and a second subset of haplotypes.

[0177] As an example of intermediate allele likelihood subsets 810a and 810b, in some embodiments, the accelerated genotype imputation system 106 sends a subset of the first round intermediate allele likelihoods from the data flow engine 802 to the accelerated computing engine 804a, so that the accelerated computing engine 804a can regenerate the first round intermediate allele likelihoods from the sacrifice round. Similarly, the accelerated genotype imputation system 106 sends an additional subset of the first round intermediate allele likelihoods from the data flow engine 802 to the accelerated computing engine 804b, so that the accelerated computing engine 804b can regenerate additional first round intermediate allele likelihoods from additional sacrifice rounds.

[0178] In addition to distributing specific data for HMM computation tasks, such as Figure 8 As further shown, the data flow engine 802 queries the HMM computing tasks of each accelerated computing engine from the accelerated computing engine cluster 804, and performs further data exchange with the on-board memory device 822. Figure 8 As shown, for example, the data flow engine 802 sends configuration and control signals 814 to the accelerated computing engine 804a, such as data indicators for the timing and order of the HMM computing tasks queued for the accelerated computing engine 804a. Similarly, in some embodiments, the data flow engine 802 receives a state signal 816 about the state or completion of a specific HMM computing task from the accelerated computing engine 804a. Based on the state signal 816 from the accelerated computing engine 804a, the data flow engine 802 queues additional HMM computing tasks for the accelerated computing engine 804a, or reorganizes or reorders other HMM computing tasks for the accelerated computing engines 804b-804n. As part of such HMM computing tasks, in some embodiments, the data flow engine 802 also receives DDR input or output requests from the onboard memory device 822 and responds to them.

[0179] As described above, in some embodiments, the accelerated genotype imputation system 106 can perform approximately 40,000 HMM calculation tasks in approximately 60 seconds, thereby speeding up processing time by 600 times. Figure 8 The configurable processor board 800 depicted in can facilitate such speeds. If the accelerated genotype imputation system 106 determines 1xα values ​​and 2xβ values ​​across a haplotype matrix of 2 trillion cells, the accelerated genotype imputation system 106 must determine the equivalent of 6 trillion cell values. Given 16 accelerated computing engines, the custom architecture in the configurable processor board 800 can perform approximately 40,000 HMM computation tasks in approximately 60 seconds.

[0180] To illustrate, if "L" represents the level of parallelism of a given accelerated computing engine (computing "L" alpha and beta values ​​per clock cycle), and the given accelerated computing engine has a core clock speed of 400mHZ, then a single accelerated computing engine can compute L units / cycle × 400M cycles per second in 60 seconds, which is equivalent to L × 24 billion alpha or beta units. In order to compute the value of 6 trillion units with 24 billion units per single accelerated computing engine, L (or the level of parallelism) would need to be equal to 16. Therefore, using Figure 8 A collection of 16 accelerated computing engines of the architecture in the configurable processor board 800 in FIG. 8 can execute approximately 40,000 HMM computation tasks in approximately 60 seconds.

[0181] In some embodiments, the accelerated computing engine may be part of a larger hardware structure. According to one or more embodiments, Fig. 9 A schematic diagram 900 of an accelerated computing engine core 914 with surrounding interfaces and other hardware is depicted.

[0182] like Fig. 9 As shown, the accelerated computing engine core 914 includes an input first-in first-out (FIFO) to receive data from the card DRAM advanced extensible interface (AXI) interface 902 and the slave address read element FIFO 912. The accelerated computing engine core 914 also includes an output FIFO to output the HMM calculation value to the write channel of the card DRAM AXI interface 902. Fig. 9 As further depicted within the accelerated computing engine core 914 in FIG. 1 , each of the input FIFO and the output FIFO includes a corresponding converter for reducing the data size and enlarging the data size, respectively.

[0183] On each side of the accelerated computing engine core 914, the schematic diagram 900 includes a buffer 910 and a buffer 916. As part of the buffer 910, the read parameter buffer and the read status buffer send or receive data from the block read state machine 920. Fig. 9As further shown, the read parameter buffer receives data from the input job FIFO 908. As part of the buffer 916, the write parameter buffer and the write status buffer send or receive data from the block write state machine 922. Additionally, the address write element FIFO 918 sends data to and receives data from the block write state machine 922 and (in some cases) the address write channel of the card DRAM AXI interface 902.

[0184] like Fig. 9 As further shown, the card DRAM AXI interface 902 includes a plurality of different channels. Specifically, the card DRAM AXI interface 902 includes an address read (AR) channel that receives data from the block read state machine 920, a write (W) channel that receives output values ​​from the accelerated computing engine core 914, and an address write (AW) channel that receives data from the block write state machine 922. In addition, the card DRAM AXI interface 902 includes a read (R) channel that receives data from the common engine wrapper (CEW) 904 and a write response (B) channel that signals response information for a write transaction.

[0185] Finally, if Fig. 9 As further shown, the CEW 904 provides access to the job control infrastructure (e.g., configuration and control signals from the data flow engine 802), the card DRAM AXI interface 902, and the host memory (e.g., the on-board memory device 822). Thus, by using the CEW 904, the accelerated genotype imputation system 106 can exchange data with the card DRAM AXI interface 902 and the streaming CEW interface 906. For example, the CEW 904 sends configuration and control signals to and from the accelerated computing engine core 914.

[0186] Now go to Fig.10 , which illustrates a flow chart of a series of actions 1000 for determining the likelihood of an intermediate allele for a genomic region including haplotype alleles by running a merge operation on a processor according to one or more embodiments of the present disclosure. Fig.10 The actions according to one embodiment are illustrated, but alternative embodiments may omit, add, reorder, and / or modify Fig.10 Any action shown. Fig.10 The actions of may be performed as part of a method. Alternatively, a non-transitory computer-readable storage medium may include a computer program that, when executed by one or more processors, causes a computing device or system to perform Fig.10In some further embodiments, the system includes at least one processor and a non-transitory computer-readable medium, which includes instructions that, when executed by the one or more processors, cause the system to perform Fig.10 instructions for actions.

[0187] like Fig.10 As shown, action 1000 includes action 1002 of identifying a haplotype reference group of a genomic region of a genomic sample. Specifically, in some embodiments, action 1002 includes identifying a haplotype reference group of a genomic region of a genomic sample using a genotype inference model. In some cases, the genotype inference model includes a hidden Markov genotype inference model.

[0188] like Fig.10 As further shown, action 1000 includes accessing the first allele possibility factor corresponding to the haplotype allele and the action 1004 corresponding to the second allele possibility factor of the haplotype allele. Specifically, in some embodiments, action 1004 includes accessing the first allele possibility factor corresponding to the haplotype allele from the haplotype reference group and the second allele possibility factor corresponding to the haplotype allele from the memory device and for the marker variant. Relatedly, in some embodiments, action 1004 includes accessing the first transition perception allele possibility factor corresponding to the haplotype allele from the haplotype reference group and the second transition perception allele possibility factor corresponding to the haplotype allele from the memory device and for the marker variant. In addition, in some cases, the memory device includes a dynamic random access memory (DRAM), a dynamic random access memory (SRAM) or a cache memory device.

[0189] For example, in some embodiments, accessing a first allele likelihood factor and a second allele likelihood factor from a memory device and for a marker variant includes accessing a first transition-aware allele likelihood factor corresponding to a haplotype allele from a haplotype reference group and a second transition-aware allele likelihood factor corresponding to a haplotype allele from a memory device and for a marker variant. In some cases, determining the first transition-aware allele likelihood factor includes combining the allele likelihood factor and the transition linear coefficient. For example, in some specific implementations, the first allele likelihood factor includes an allele likelihood factor of a sample reference haplotype allele or a sample replacement haplotype allele; and the second allele likelihood factor includes an allele likelihood factor of a sample reference haplotype allele or a sample replacement haplotype allele.

[0190] Relatedly, in some embodiments, act 1000 also includes predetermining a first transition-aware allele likelihood factor and a second transition-aware allele likelihood factor as part of a round across the haplotype matrix before determining one or more intermediate allele likelihoods corresponding to the marker variants. Similarly, in some cases, act 1000 includes predetermining a first transition-aware allele likelihood factor and a second transition-aware allele likelihood factor before determining one or more intermediate allele likelihoods corresponding to the marker variants. For example, in some embodiments, act 1004 includes predetermining a first transition-aware allele likelihood factor by combining an allele likelihood factor for a haplotype allele and a transition constant coefficient for transitioning between haplotypes from a haplotype reference group; and predetermining a second transition-aware allele likelihood factor by combining an allele likelihood factor and a transition linear coefficient for transitioning between haplotypes from a haplotype reference group.

[0191] like Fig.10 As further shown, act 1000 includes combining the first allele likelihood factor and the neighboring marker intermediate allele likelihood to generate a neighboring marker factor-aware allele likelihood act 1006. Specifically, in some implementations, act 1006 includes combining the first allele likelihood factor and the neighboring marker intermediate allele likelihood for a genomic region including a haplotype allele given a neighboring marker variant to generate a neighboring marker factor-aware allele likelihood for the marker variant and a haplotype from a haplotype reference set.

[0192] In addition, in some cases, action 1006 includes combining, by a configurable processor, a first transition-aware allele likelihood factor for a genomic region including a haplotype allele and a neighboring marker intermediate allele likelihood given a neighboring marker variant to generate a neighboring marker transition factor-aware allele likelihood for the marker variant and a haplotype from a haplotype reference set. For example, in some embodiments, the configurable processor includes an application specific integrated circuit (ASIC), an application specific standard product (ASSP), a coarse-grained reconfigurable array (CGRA), or a field programmable gate array (FPGA).

[0193] To further illustrate, in some embodiments, combining the first allele likelihood factor and the neighboring marker intermediate allele likelihood comprises multiplying the first transition-aware allele likelihood factor and the neighboring marker intermediate allele likelihood without further multiplication to determine the intermediate allele likelihood. Relatedly, in certain specific implementations, combining the first transition-aware allele likelihood factor and the neighboring marker intermediate allele likelihood comprises multiplying the first transition-aware allele likelihood factor and the neighboring marker intermediate allele likelihood without further multiplication to determine the intermediate allele likelihood.

[0194] like Fig.10 As further shown, action 1000 includes action 1008 of determining an intermediate allele probability based on the adjacent marker factor sensed allele probability and the second allele probability factor. Specifically, in some specific implementations, action 1008 includes determining, for marker variants and haplotypes, an intermediate allele probability for a genomic region including haplotype alleles based on the adjacent marker factor sensed allele probability and the second allele probability factor. In addition, in some cases, action 1008 includes determining, by a configurable processor and for marker variants and haplotypes, an intermediate allele probability for a genomic region including haplotype alleles based on the adjacent marker transition factor sensed allele probability and the second transition sensed allele probability factor.

[0195] In addition, in some cases, determining the intermediate allele likelihood includes determining the intermediate allele likelihood for a genomic region including the sample reference haplotype allele or the sample alternative haplotype allele. Relatedly, in some cases, determining the intermediate allele likelihood based on the neighboring marker factor perceived allele likelihood and the second allele likelihood factor includes summing the neighboring marker transition factor perceived allele likelihood and the summed neighboring marker transition perceived allele likelihood factor.

[0196] like Fig.10 As further shown, action 1000 includes an action 1010 to generate an allele probability based on an intermediate allele probability. Specifically, in some implementations, action 1008 includes, for a set of marker variants corresponding to a genomic region, generating an allele probability for a genomic region including haplotype alleles from a haplotype reference group based on the intermediate allele probability. In addition, in some cases, action 1010 includes, by a configurable processor and for a set of marker variants corresponding to a genomic region, generating an allele probability for a genomic region including haplotype alleles from a haplotype reference group based on the intermediate allele probability.

[0197] As a supplement or alternative to actions 1002-1010, in some specific implementations, action 1000 also includes sending a corresponding input value set including an allele likelihood factor, a transition coefficient, and a haplotype allele value from the data flow engine to a corresponding accelerated computing engine in the accelerated computing engine cluster; and determining, by the corresponding accelerated computing engine and based on the corresponding input value set, a corresponding intermediate allele likelihood set corresponding to a corresponding subset of marker variants and a corresponding subset of haplotypes. In some embodiments, the data flow engine corresponds to the accelerated computing engine cluster.

[0198] To further illustrate, in some cases, action 1000 also includes sending a corresponding input value set from the data flow engine to a corresponding accelerated computing engine by the following steps: sending a first input value set including allele probability factors, transition coefficients, and haplotype allele values ​​from the data flow engine to a first accelerated computing engine in the accelerated computing engine cluster; sending a second input value set including allele probability factors, transition coefficients, and haplotype allele values ​​from the data flow engine to a second accelerated computing engine in the accelerated computing engine cluster; and determining a corresponding intermediate allele probability set by the following steps: determining a first intermediate allele probability set corresponding to a first subset of marker variants and a first subset of haplotypes by the first accelerated computing engine and based on the first input value set; and determining a second intermediate allele probability set corresponding to a second subset of marker variants and a second subset of haplotypes by the second accelerated computing engine and based on the second input value set.

[0199] As suggested above, in some cases, action 1000 also includes: accessing a second transition-aware allele likelihood factor as part of a summed neighboring marker transition-aware allele likelihood factor; and determining an intermediate allele likelihood based on the neighboring marker transition factor-aware allele likelihood and the summed neighboring marker transition-aware allele likelihood factor. Relatedly, in some specific implementations, action 1000 includes predetermining the summed neighboring marker transition-aware allele likelihood factor by combining the allele likelihood factor of a haplotype allele, a transition constant coefficient for transitioning between haplotypes from a haplotype reference group, and the summed neighboring marker intermediate allele likelihood of neighboring marker variants. As further suggested above, in some cases, the allele likelihood factor of a haplotype allele includes a reference allele likelihood factor of a sample reference haplotype allele or an alternative allele likelihood factor of a sample alternative haplotype allele.

[0200] Additionally, in some implementations, act 1000 also includes determining one or more nucleobase calls for a genomic region from a genomic sample based on the allele likelihood of the genomic region and one or more variant nucleobase calls surrounding the genomic region.

[0201] Now go to Fig.11 , which illustrates a flow chart of a series of actions 1100 for determining and storing a subset of intermediate allele probabilities corresponding to a marker variant group, and temporarily generating an intermediate allele probability set for a marker variant set by using the intermediate allele probability subset as a hot start point according to one or more embodiments of the present disclosure. Fig.11 The actions according to one embodiment are illustrated, but alternative embodiments may omit, add, reorder, and / or modify Fig.11 Any action shown. Fig.11 The actions of may be performed as part of a method. Alternatively, a non-transitory computer-readable storage medium may include a computer program that, when executed by one or more processors, causes a computing device or system to perform Fig.11 In some further embodiments, the system includes at least one processor and a non-transitory computer-readable medium, which includes instructions that, when executed by the one or more processors, cause the system to perform Fig.11 instructions for actions.

[0202] like Fig.11 As shown, action 1100 includes action 1102 of determining the first round of intermediate allele possibilities. Specifically, in some embodiments, action 1102 includes determining the first round of intermediate allele possibilities of a genomic region from a genomic sample including haplotype alleles corresponding to a haplotype set under the condition of a given marker variant set by performing a first round. In addition, in some cases, action 1102 includes determining the first round of intermediate allele possibilities of a genomic region from a genomic sample including haplotype alleles corresponding to a haplotype set under the condition of a given marker variant set using a configurable processor that performs the first round. In some cases, the configurable processor includes an application specific integrated circuit (ASIC), an application specific standard product (ASSP), a coarse-grained reconfigurable array (CGRA), or a field programmable gate array (FPGA).

[0203] like Fig.11 As further shown, action 1100 includes an action 1104 storing a subset of the first round of intermediate allele possibilities. Specifically, in some embodiments, action 1104 includes storing a subset of the first round of intermediate allele possibilities corresponding to a subset of the marker variants of the marker variant group on a memory device. In addition, in some cases, action 1104 includes storing a subset of the first round of intermediate allele possibilities corresponding to a subset of the marker variants of the marker variant group. In some cases, the memory device includes a dynamic random access memory (DRAM), a dynamic random access memory (SRAM) or a cache memory device.

[0204] like Fig.11 As further shown in, action 1100 includes an action 1106 of regenerating the first round of intermediate allele possibilities based on a subset of the first round of stored intermediate allele possibilities. Specifically, in some specific implementations, action 1106 includes initializing the allele possibility determination at the marker variant group by utilizing the stored subset of the first round of intermediate allele possibilities to regenerate the first round of intermediate allele possibilities. In addition, in some embodiments, action 1106 includes utilizing a configurable processor to initializing the allele possibility determination at the marker variant group by utilizing the stored subset of the first round of intermediate allele possibilities to regenerate the first round of intermediate allele possibilities.

[0205] Relatedly, in some cases, initializing allele probability determination at a marker variant group using a subset of stored first-round intermediate allele probabilities includes: determining a first subset of first-round intermediate allele probabilities for the first marker variant group based on a first stored column of first-round intermediate allele probabilities for initial marker variants from the first marker variant group; and determining a second subset of first-round intermediate allele probabilities for the second marker variant group based on a second stored column of first-round intermediate allele probabilities for initial marker variants from the second marker variant group.

[0206] Relatedly, in some cases, action 1100 includes storing a subset of first-round intermediate allele probabilities by: storing the subset of first-round intermediate allele probabilities in a dynamic random access memory (DRAM); and initializing allele probability determination at a marker variant group using the stored subset of first-round intermediate allele probabilities includes accessing the stored subset of first-round intermediate allele probabilities from the DRAM.

[0207] like Fig.11 As further shown, act 1100 includes an act 1108 of determining a second round of intermediate allele likelihoods. Specifically, in some implementations, act 1108 includes determining, by performing the second round, a second round of intermediate allele likelihoods for genomic regions that include haplotype alleles corresponding to the haplotype set given the set of marker variants. Additionally, in some cases, act 1108 includes determining, using a configurable processor that performs the second round, a second round of intermediate allele likelihoods for genomic regions that include haplotype alleles corresponding to the haplotype set given the set of marker variants.

[0208] As suggested above, in some cases, action 1100 includes: determining a first round intermediate allele probability includes using a reverse round to determine a reverse intermediate allele probability for a genomic region including a haplotype allele; and determining a second round intermediate allele probability includes using a forward round to determine a forward intermediate allele probability for a genomic region including a haplotype allele.

[0209] like Fig.11 As further shown, action 1100 includes an action 1110 of generating allele probabilities based on the regenerated first round intermediate allele probabilities and the second round intermediate allele probabilities. Specifically, in some implementations, action 1110 includes generating allele probabilities for genomic regions including haplotype alleles based on the regenerated first round intermediate allele probabilities and the second round intermediate allele probabilities. In addition, in some embodiments, action 1110 includes generating allele probabilities for genomic regions including haplotype alleles based on the regenerated first round intermediate allele probabilities and the second round intermediate allele probabilities using an output engine.

[0210] To illustrate, in some embodiments, generating allele probabilities based on regenerated first-round intermediate allele probabilities and second-round intermediate allele probabilities includes: determining a summed first-round intermediate allele probability for a set of marker variants based on the regenerated first-round intermediate allele probability; determining a summed second-round intermediate allele probability for a set of marker variants based on the second-round intermediate allele probability; and determining the allele probability based on the summed first-round intermediate allele probability and the summed second-round intermediate allele probability.

[0211] In addition or in lieu of actions 1102-1110, in some implementations, action 1000 further includes: storing haplotype allele indicator data in a haplotype allele indicator memory; storing transition coefficients in a transition coefficient memory; and storing allele likelihood factors in an allele likelihood factor memory. In addition, in some embodiments, action 1000 includes using a joint engine to determine an intermediate allele likelihood value.

[0212] As suggested above, in some cases, action 1100 also includes: sending a corresponding input value set including allele likelihood factors, transition coefficients, and haplotype allele values ​​from the data flow engine to a corresponding accelerated computing engine in the accelerated computing engine cluster; and determining, by the corresponding accelerated computing engine and based on the corresponding input value set, a corresponding intermediate allele likelihood set corresponding to a corresponding subset of marker variants and a corresponding subset of haplotypes. In some embodiments, the data flow engine corresponds to the accelerated computing engine cluster.

[0213] To further illustrate, in some cases, action 1100 also includes sending a corresponding input value set from the data flow engine to a corresponding accelerated computing engine by the following steps: sending a first input value set including allele probability factors, transition coefficients, and haplotype allele values ​​from the data flow engine to a first accelerated computing engine in the accelerated computing engine cluster; sending a second input value set including allele probability factors, transition coefficients, and haplotype allele values ​​from the data flow engine to a second accelerated computing engine in the accelerated computing engine cluster; and determining a corresponding intermediate allele probability set by the following steps: determining a first intermediate allele probability set corresponding to a first subset of marker variants and a first subset of haplotypes by the first accelerated computing engine and based on the first input value set; and determining a second intermediate allele probability set corresponding to a second subset of marker variants and a second subset of haplotypes by the second accelerated computing engine and based on the second input value set.

[0214] As further suggested above, in some cases, action 1100 includes: sending a subset of the first-round intermediate allele probabilities from the data flow engine to a first accelerated computing engine in the accelerated computing engine cluster for the first accelerated computing engine to regenerate the first-round intermediate allele probabilities; and sending an additional subset of the first-round intermediate allele probabilities from the data flow engine to a second accelerated computing engine in the accelerated computing engine cluster for the second accelerated computing engine to regenerate additional first-round intermediate allele probabilities.

[0215] Furthermore, in some specific implementations, action 1100 includes: sending a subset of the first round intermediate allele likelihoods from a memory device to a data flow engine; and sending a subset of the first round intermediate allele likelihoods from the data flow engine to an accelerated computing engine to regenerate the first round intermediate allele likelihoods based on the subset of the first round intermediate allele likelihoods. Additionally, in some cases, action 1100 includes: storing haplotype allele indicator data of a haplotype matrix on a memory device; and accessing haplotype allele indicator data of a haplotype matrix from a memory device to generate allele likelihoods using a hidden Markov haploid genotype inference model or a hidden Markov diploid genotype inference model.

[0216] As suggested above, in some cases, act 1100 includes determining one or more nucleobase calls for a genomic region from a genomic sample based on the allelic likelihood of the genomic region and one or more variant nucleobase calls surrounding the genomic region.

[0217] As further suggested above, in certain embodiments, action 1100 includes: storing haplotype allele indicator data of the haplotype matrix on a dynamic random access memory (DRAM); and accessing the haplotype allele indicator data of the haplotype matrix from the DRAM by a configurable processor to generate allele probabilities using a hidden Markov haploid genotype inference model or a hidden Markov diploid genotype inference model.

[0218] Additionally or alternatively, in some embodiments, action 1100 includes: determining a running sum of a first subset of intermediate allele probabilities for genomic regions including haplotype alleles of a first type of one or more haplotypes from a haplotype reference group for neighboring marker variants; determining a running sum of a second subset of intermediate allele probabilities for genomic regions including haplotype alleles of a second type from one or more haplotypes for neighboring marker variants; and determining a sum of intermediate allele probabilities for genomic regions including haplotype alleles of haplotypes from a haplotype reference group for marker variants based on the running sum of the first subset of intermediate allele probabilities and the running sum of the second subset of intermediate allele probabilities.

[0219] Now go to Fig.12 , which illustrates a flow chart of a series of actions 1200 for determining a running sum of the median allele likelihoods for genomic regions that exhibit haplotype alleles for one or more haplotypes given one marker variant, and using the running sum as a running input to determine the median allele likelihoods for genomic regions that exhibit haplotype alleles given another marker variant, according to one or more embodiments of the present disclosure. Fig.12 The actions according to one embodiment are illustrated, but alternative embodiments may omit, add, reorder, and / or modify Fig.12 Any action shown. Fig.12 The actions of may be performed as part of a method. Alternatively, a non-transitory computer-readable storage medium may include a computer program that, when executed by one or more processors, causes a computing device or system to perform Fig.12 In some further embodiments, the system includes at least one processor and a non-transitory computer-readable medium, which includes instructions that, when executed by the one or more processors, cause the system to perform Fig.12 instructions for actions.

[0220] like Fig.12 As shown, action 1200 includes action 1202 of identifying a haplotype reference group of a genomic region of a genomic sample. Specifically, in some embodiments, action 1202 includes identifying a haplotype reference group of a genomic region of a genomic sample using a genotype inference model.

[0221] like Fig.12 As further shown, act 1200 includes determining a running sum of a first subset of intermediate allele probabilities for neighboring marker variants 1204. Specifically, in some embodiments, act 1204 includes determining a running sum of a first subset of intermediate allele probabilities for neighboring marker variants for genomic regions that include haplotype alleles of a first type of one or more haplotypes from a haplotype reference set.

[0222] like Fig.12 As further shown, act 1200 includes determining a running sum of a second subset of intermediate allele probabilities for the adjacent marker variants 1206. Specifically, in some implementations, act 1206 includes determining a running sum of a second subset of intermediate allele probabilities for the adjacent marker variants for genomic regions including haplotype alleles of a second type from one or more haplotypes.

[0223] As described above, in some embodiments, the first type of haplotype alleles comprises sample reference haplotype alleles and the second type of haplotype alleles comprises sample alternative haplotype alleles.

[0224] like Fig.12 As further shown, act 1200 includes determining a sum of intermediate allele probabilities for the marker variants based on a running sum of the first subset of intermediate allele probabilities and a running sum of the second subset of intermediate allele probabilities act 1208. Specifically, in some implementations, act 1208 includes determining a sum of intermediate allele probabilities for the genomic region including haplotype alleles of a haplotype from a haplotype reference set based on a running sum of the first subset of intermediate allele probabilities and a running sum of the second subset of intermediate allele probabilities for the marker variants.

[0225] As indicated above, in some cases, determining the sum of intermediate allele probabilities includes determining, by a configurable processor and for marker variants, an initial intermediate allele probability based on intermediate allele probabilities from a first subset of intermediate allele probabilities or from a second subset of intermediate allele probabilities and from the intermediate allele probabilities before summing neighboring marker intermediate allele probabilities for genomic regions including haplotype alleles for neighboring marker variants.

[0226] Additionally or alternatively, in some embodiments, determining the sum of intermediate allele probabilities includes determining, by a configurable processor and for a marker variant, an initial intermediate allele probability based on an intermediate allele probability from a first subset of intermediate allele probabilities or from a second subset of intermediate allele probabilities and from the intermediate allele probabilities before generating allele probabilities for a genomic region including haplotype alleles for adjacent marker variants.

[0227] like Fig.12 As further shown, act 1200 includes generating an allele likelihood based on the sum of the intermediate allele likelihoods 1210. Specifically, in some implementations, act 1210 includes generating an allele likelihood for a genomic region including haplotype alleles based on the sum of the intermediate allele likelihoods.

[0228] In addition to or in lieu of actions 1202-1210, in some embodiments, action 1000 also includes: predetermining a first transition-aware allele likelihood factor for rows corresponding to the first type of haplotype alleles and a second transition-aware allele likelihood factor for rows corresponding to the second type of haplotype alleles; and further determining the sum of intermediate allele probabilities based on the first transition-aware allele likelihood factor for rows corresponding to the first type of haplotype alleles and the second transition-aware allele likelihood factor for rows corresponding to the second type of haplotype alleles.

[0229] Relatedly, in some cases, a neighboring marker sum of intermediate allele probabilities for a genomic region including the haplotype allele is determined for a neighboring marker variant; and a sum of intermediate allele probabilities is further determined for a neighboring marker variant based on a combination of the neighboring marker sum of intermediate allele probabilities, a first transition-aware allele probability factor for a row corresponding to a first type of haplotype allele, and a second transition-aware allele probability factor for a row corresponding to a second type of haplotype allele.

[0230] As suggested above, in some cases, action 1200 also includes: multiplying a running sum of a first subset of intermediate allele probabilities by a first transition-aware allele probability factor; multiplying a running sum of a second subset of intermediate allele probabilities by a second transition-aware allele probability factor; and determining a sum of intermediate allele probabilities for marker variants based on the multiplied running sum of the first subset of intermediate allele probabilities and the multiplied running sum of the second subset of intermediate allele probabilities.

[0231] As further suggested above, in some embodiments, action 1200 includes: predetermining a first transition-aware allele likelihood factor comprises combining a first allele likelihood factor for a haplotype allele of a first type and a transition linear coefficient for transitioning between haplotypes from a haplotype reference group; and predetermining a second transition-aware allele likelihood factor comprises combining a second allele likelihood factor for a haplotype allele of a second type and the transition linear coefficient.

[0232] Methods described herein can be used in combination with multiple nucleic acid sequencing techniques. Particularly suitable techniques are those in which nucleic acids are attached to fixed positions in an array so that their relative positions do not change and in which the array is repeatedly imaged. Embodiments in which images are obtained in different color channels (e.g., matching different markers for distinguishing a type of core base from another type of core base) are particularly suitable. In some embodiments, the process of determining the nucleotide sequence of a target nucleic acid (i.e., nucleic acid polymer) can be an automated process. Preferred embodiments include sequencing by synthesis (SBS) technology.

[0233] SBS techniques generally include enzymatic extension of nascent nucleic acid chains by repeated addition of nucleotides to the template strand. In traditional SBS methods, a single nucleotide monomer can be provided to the target nucleotide in the presence of a polymerase in each delivery. However, in the methods described herein, more than one type of nucleotide monomer can be provided to the target nucleic acid in the presence of a polymerase in the delivery.

[0234] SBS can utilize nucleotide monomers with terminator parts or lack nucleotide monomers of any terminator parts.The method utilizing the nucleotide monomers lacking terminator includes for example pyrophosphate sequencing and the sequencing of nucleotides using gamma-phosphate labeling, as described in further detail below.In the method using the nucleotide monomers lacking terminator, the number of nucleotides added in each cycle is usually variable, and the number depends on the mode of template sequence and nucleotide delivery.For utilizing the SBS technology of the nucleotide monomers with terminator parts, terminator can be effectively irreversible under the sequencing conditions used, such as the situation of traditional Sanger sequencing utilizing dideoxynucleotides, or terminator can be reversible, such as the situation of the sequencing method developed by Solexa (now Illumina, Inc.).

[0235] The SBS technique can utilize nucleotide monomers with a labeling portion or nucleotide monomers lacking a labeling portion. Thus, incorporation events can be detected based on the following items: properties of the label, such as the fluorescence of the label; properties of the nucleotide monomer, such as molecular weight or charge; byproducts of the incorporated nucleotide, such as the release of pyrophosphate; and the like. In embodiments where two or more different nucleotides are present in the sequencing reagent, the different nucleotides may be distinguishable from each other, or alternatively, the two or more different labels may be indistinguishable under the detection technology used. For example, different nucleotides present in the sequencing reagent may have different labels, and they may be distinguished using appropriate optical devices, as exemplified by the sequencing method developed by Solexa (now Illumina, Inc.).

[0236] Preferred embodiments include pyrophosphate sequencing technology. Pyrophosphate sequencing detects the release of inorganic pyrophosphate (PPi) when a specific nucleotide is incorporated into a nascent chain (Ronaghi, M., Karamohamed, S., Pettersson, B., Uhlen, M., and Nyren, P. (1996), "Real-time DNA sequencing using detection of pyrophosphate release.", Analytical Biochemistry 242 (1), 84-9; Ronaghi, M. (2001), "Pyrosequencing sheds light on DNA sequencing.", Genome Res. 11 (1), 3-11; Ronaghi, M., Uhlen, M., and Nyren, P. (1998), "A sequencing method based on real-time pyrophosphate.", Science 281 (5375), 363; U.S. Pat. No. 6,210,891; U.S. Pat. No. 6,258,568 and U.S. Pat. No. 6,274,320, the disclosures of which are incorporated herein by reference in their entirety). In pyrophosphate sequencing, the released PPi can be detected by being immediately converted into ATP by adenosine triphosphate (ATP) sulfurylase, and the level of ATP generated is detected by photons generated by luciferase. The nucleic acid to be sequenced can be attached to a feature in an array, and the array can be imaged to capture the chemiluminescent signal generated by the incorporation of nucleotides at the feature of the array. The image can be obtained after the array is treated with a specific nucleotide type (e.g., A, T, C or G). The images obtained after adding each nucleotide type will differ in which features in the array are detected. These differences in the image reflect the different sequence contents of the features on the array. However, the relative position of each feature will remain unchanged in the image. The images can be stored, processed and analyzed using the methods described herein. For example, images obtained after treating the array with each different nucleotide type can be processed in the same manner as exemplified herein for images obtained from different detection channels for a reversible terminator-based sequencing method.

[0237] In another exemplary type of SBS, cycle sequencing is accomplished by the stepwise addition of reversible terminator nucleotides, which contain, for example, cleavable or photobleachable dye labels, as described in, for example, WO 04 / 018497 and U.S. Patent No. 7,057,026, the disclosures of which are incorporated herein by reference. The method is commercialized by Solexa (now Illumina Inc.) and is also described in WO 91 / 06678 and WO 07 / 123,744, the disclosures of each of which are incorporated herein by reference. The availability of fluorescently labeled terminators, in which termination can be reversible and the fluorescent label can be cleaved, facilitates efficient cyclic reversible termination (CRT) sequencing. Polymerases can also be co-engineered to efficiently incorporate these modified nucleotides and extend from these modified nucleotides.

[0238] Preferably, in the sequencing embodiment based on reversible terminator, the label does not substantially inhibit extension under SBS reaction conditions. However, the detection label can be removable, such as removed by cleavage or degradation. Images can be captured after the label is incorporated into the arrayed nucleic acid feature. In a specific embodiment, each cycle involves four different nucleotide types being delivered to the array simultaneously, and each nucleotide type has a label that is spectrally different. Four images can then be obtained, each image using a detection channel that is selective to one label in the four different labels. Alternatively, different nucleotide types can be added sequentially, and the image of the array can be obtained between each addition step. In such embodiments, each image will show the nucleic acid feature that has been incorporated with a specific type of nucleotide. Due to the different sequence content of each feature, different feature portions are present or absent in different images. However, the relative position of the feature portion will remain unchanged in the image. The image obtained by such reversible terminator-SBS method can be stored, processed and analyzed as described herein. After the image capture step, the label can be removed and the reversible terminator portion can be removed for subsequent cycles of nucleotide addition and detection. Removing labels after they have been detected in a particular cycle and before subsequent cycles can provide the advantage of reducing background signal and crosstalk between cycles. Examples of useful labeling and removal methods are set forth below.

[0239] In a specific embodiment, some or all of the nucleotide monomers may include a reversible terminator. In such embodiments, the reversible terminator / cleavable fluorophore may include a fluorophore (Metzker, Genome Res. 15: 1767-1776 (2005), which is incorporated herein by reference) connected to a ribose moiety via a 3' ester bond. Other methods have separated terminator chemistry from the cleavage of a fluorescent label (Ruparel et al., Proc Natl Acad Sci USA 102: 5932-7 (2005), which is incorporated herein by reference in its entirety). Ruparel et al. describe the development of reversible terminators that use small 3' allyl groups to block extension, but can be easily deblocked by a short treatment with a palladium catalyst. The fluorophore is attached to the base via a photocleavable linker that can be easily cleaved by exposure to long wavelength ultraviolet light for 30 seconds. Therefore, disulfide reduction or photocleavage can be used as a cleavable linker. Another method of reversible termination is to use natural termination, which occurs after a bulky dye is placed on a dNTP. The presence of a charged bulky dye on a dNTP can act as an efficient terminator by steric hindrance and / or electrostatic hindrance. Unless the dye is removed, the presence of an incorporation event prevents further incorporation. The cracking of the dye removes the fluorophore and effectively reverses termination. The example of modified nucleotides is also described in U.S. Patent No. 7,427,673 and U.S. Patent No. 7,057,026, the disclosure of which is incorporated herein by reference in its entirety.

[0240] Additional exemplary SBS systems and methods that may be utilized with the methods and systems described herein are described in U.S. Patent Application Publication No. 2007 / 0166705, U.S. Patent Application Publication No. 2006 / 0188901, U.S. Patent No. 7,057,026, U.S. Patent Application Publication No. 2006 / 0240439, U.S. Patent Application Publication No. 2006 / 0281109, PCT Publication No. WO05 / 065814, U.S. Patent Application Publication No. 2005 / 0100900, PCT Publication No. WO 06 / 064199, PCT Publication No. WO07 / 010,251, U.S. Patent Application Publication No. 2012 / 0270305, and U.S. Patent Application Publication No. 2013 / 0260372, the disclosures of which are incorporated herein by reference in their entireties.

[0241] Some embodiments can use the detection of four different nucleotides using less than four different marks.For example, the method and system described in the material of the U.S. Patent Application Publication No. 2013 / 0079232 incorporated can be used to perform SBS.As a first example, a pair of nucleotide types can be detected at the same wavelength, but based on the intensity difference of a member in the pair relative to another member, or based on a member in the pair causing a change (for example, by chemical modification, photochemical modification or physical modification) that a signal that is obvious compared with the signal of another member of the pair detected appears or disappears.As a second example, three of the four different nucleotide types can be detected under specific conditions, and the fourth nucleotide type lacks a mark (for example, the minimum detection caused by background fluorescence, etc.) that can be detected under those conditions or detected minimally under those conditions.The first three nucleotide types can be determined to be incorporated into nucleic acid based on the existence of its corresponding signal, and the fourth nucleotide type can be determined to be incorporated into nucleic acid based on the absence of any signal or the minimum detection of any signal.As a third example, a nucleotide type can include a mark detected in two different channels, and other nucleotide types are detected in no more than one channel. The three exemplary configurations described above are not considered mutually exclusive and can be used in various combinations. An exemplary embodiment combining all three examples is a fluorescence-based SBS method that uses a first nucleotide type detected in a first channel (e.g., dATP with a label detected in the first channel when excited by a first excitation wavelength), a second nucleotide type detected in a second channel (e.g., dCTP with a label detected in the second channel when excited by a second excitation wavelength), a third nucleotide type detected in both the first channel and the second channel (e.g., dTTP with at least one label detected in both channels when excited by the first excitation wavelength and / or the second excitation wavelength), and a fourth nucleotide type that lacks a label detected or minimally detected in either channel (e.g., dGTP without a label).

[0242] In addition, as described in the materials of incorporated U.S. Patent Application Publication No. 2013 / 0079232, a single channel can be used to obtain sequencing data. In such so-called single dye sequencing methods, a first nucleotide type is marked, but the mark is removed after the first image is generated, and a second nucleotide type is marked only after the first image is generated. A third nucleotide type retains its mark in both the first image and the second image, and a fourth nucleotide type remains unmarked in both images.

[0243] Some embodiments can utilize sequencing technology while connecting. Such technology utilizes DNA ligase to incorporate oligonucleotides and recognize the incorporation of such oligonucleotides. Oligonucleotides generally have different labels related to the identity of specific nucleotides in the sequence of oligonucleotide hybridization. As with other SBS methods, images can be obtained after the array of nucleic acid features is treated with labeled sequencing reagents. Each image will show the nucleic acid features that have been incorporated with a specific type of label. Due to the different sequence content of each feature, different features are present or absent in different images, but the relative position of the features will remain unchanged in the image. The images obtained by the sequencing method based on connection can be stored, processed and analyzed as described herein. Exemplary SBS systems and methods that can be used with the methods and systems described herein are described in U.S. Patent No. 6,969,488, U.S. Patent No. 6,172,218 and U.S. Patent No. 6,306,597, and the disclosures of these patents are incorporated herein by reference in their entirety.

[0244] Some embodiments may utilize nanopore sequencing (Deamer, DW and Akeson, M., "Nanopores and nucleic acids: prospects for ultrarapid sequencing.", Trends Biotechnol. 18, 147-151 (2000); Deamer, D. and D. Branton, "Characterization of nucleic acids by nanopore analysis.", Acc. Chem. Res. 35: 817-825 (2002); Li, J., M. Gershow, D. Stein, E. Brandin and JA Golovchenko, "DNA molecules and configurations in a solid-state nanopore microscope", Nat. Mater., 2: 611-615 (2003), the disclosures of which are incorporated herein by reference in their entirety). In such embodiments, the target nucleic acid passes through the nanopore. The nanopore may be a synthetic pore or a biological membrane protein, such as α-hemolysin. As the target nucleic acid passes through the nanopore, each base pair can be identified by measuring the fluctuations in the conductivity of the pore. (U.S. Pat. No. 7,001,792; Soni, GV and Meller, "A. Progress toward ultrafast DNA sequencing using solid-state nanopores.", Clin. Chem. 53, 1996-2001 (2007); Healy, K., "Nanopore-based single-molecule DNA analysis.", Nanomed., 2, 459-481 (2007); Cockroft, SL, Chu, J., Amorin, M. and Ghadiri, MR, "A single-molecule nanopore device detects DNA polymerase activity with single-nucleotide resolution.", J. Am. Chem. Soc. 130, 818-820 (2008), the disclosures of which are incorporated herein by reference in their entirety). The data obtained from nanopore sequencing can be stored, processed and analyzed as described herein. In particular, according to the exemplary processing of optical images and other images described herein, the data can be processed like an image.

[0245] Some embodiments may utilize methods involving real-time monitoring of DNA polymerase activity. Nucleotide incorporation may be detected by fluorescence resonance energy transfer (FRET) interactions between a polymerase carrying a fluorophore and a γ-phosphate labeled nucleotide, as described, for example, in U.S. Pat. No. 7,329,492 and U.S. Pat. No. 7,211,414 (each of which is incorporated herein by reference), or may be detected by zero-mode waveguides, as described, for example, in U.S. Pat. No. 7,315,019 (which is incorporated herein by reference), and may be detected by fluorescent nucleotide analogs and engineered polymerases, as described, for example, in U.S. Pat. No. 7,405,281 and U.S. Patent Application Publication No. 2008 / 0108082 (each of which is incorporated herein by reference). Illumination can be limited to a volume of the order of magnitude surrounding the surface-tethered polymerase, so that the incorporation of fluorescently labeled nucleotides can be observed under low background (Levene, MJ et al., "Zero-mode waveguides for single-molecule analysis at high concentrations.", Science 299, 682-686 (2003); Lundquist, PM et al., "Parallel confocal detection of single molecules in real time.", Opt. Lett. 33, 1026-1028 (2008); Korlach, J. et al., "Selective aluminum passivation for targeted immobilization of single DNA polymerase molecules in zero-mode waveguide nano structures.", Proc. Natl. Acad. Sci. USA 105, 1176-1181 (2008), the disclosures of which are incorporated herein by reference in their entirety). Images obtained by such methods can be stored, processed, and analyzed as described herein.

[0246] Some SBS embodiments include detecting the proton released when nucleotides are incorporated into extension products. For example, sequencing based on the detection of releasing protons can use the commercially available electrical detectors and related technologies available from Ion Torrent company (Guilford, CT, it is a Life Technologies subsidiary) or the sequencing methods and systems described in US 2009 / 0026082A1, US2009 / 0127589 A1, US2010 / 0137143 A1 or US 2010 / 0282617A1, each of which is incorporated herein by reference. The method for amplifying target nucleic acids using kinetic exclusion set forth herein can be easily applied to substrates for detecting protons. More specifically, the method set forth herein can be used to produce amplicon clone colonies for detecting protons.

[0247] The above-mentioned SBS method can be advantageously carried out in a variety of formats so that a plurality of different target nucleic acids are manipulated simultaneously. In a specific embodiment, different target nucleic acids can be processed in a common reaction vessel or on the surface of a specific substrate. This allows to conveniently deliver sequencing reagents, remove unreacted reagents and detect incorporation events in a variety of ways. In the embodiment using the target nucleic acid of surface binding, the target nucleic acid can be an array format. In the array format, the target nucleic acid can be attached to the surface in a spatially distinguishable manner generally. The target nucleic acid can be attached by direct covalent attachment, attached to beads or other particles or attached to a polymerase or other molecules attached to the surface to combine. The array can include a single copy of the target nucleic acid at each site (also referred to as a feature portion), or multiple copies with the same sequence can be present at each site or feature portion. Multiple copies can be produced by an amplification method (such as bridge amplification or emulsion PCR as described in further detail below).

[0248] The methods described herein may use arrays having features at any of a variety of densities, including, for example, at least about 10 features / cm2, 100 features / cm2, 500 features / cm2, 1,000 features / cm2, 5,000 features / cm2, 10,000 features / cm2, 50,000 features / cm2, 100,000 features / cm2, 1,000,000 features / cm2, 5,000,000 features / cm2, or more.

[0249] The advantage of the method described herein is that they provide rapid and effective detection of multiple target nucleic acids in parallel. Therefore, the present disclosure provides an integrated system that can use technology known in the art (such as those illustrated above) to prepare and detect nucleic acids. Therefore, the integrated system of the present disclosure may include a fluid assembly that can deliver amplification reagents and / or sequencing reagents to one or more fixed DNA fragments, and the system includes components such as pumps, valves, reservoirs, fluid pipelines, etc. The circulation cell can be configured for and / or for detecting target nucleic acids in the integrated system. Exemplary circulation cells are described in, for example, US 2010 / 0111768A1 and U.S. Serial No. 13 / 273,666, each of which is incorporated herein by reference. As illustrated for the circulation cell, one or more fluid assemblies of the integrated system can be used for amplification method and detection method. Taking nucleic acid sequencing embodiment as an example, one or more fluid assemblies of the integrated system can be used for the amplification method described herein and for delivering sequencing reagents in sequencing method (such as those illustrated above). Alternatively, the integrated system may include a separate fluid system to perform amplification method and perform detection method. Examples of integrated sequencing systems capable of producing amplified nucleic acids and also determining nucleic acid sequences include, but are not limited to, MiSeq TM The ELISA platform (Illumina, Inc., San Diego, CA) and the apparatus described in U.S. Ser. No. 13 / 273,666, which is incorporated herein by reference.

[0250] The above-mentioned sequencing system sequences the nucleic acid polymers present in the sample received by the sequencing device. As defined herein, "sample" and its derivatives are used in their broadest sense, including any specimens, cultures, etc. suspected of containing a target. In some embodiments, the sample includes DNA, RNA, PNA, LNA, chimeric or hybridized nucleic acids. The sample may include any biological, clinical, surgical, agricultural, atmospheric or aquatic animal and plant specimens containing one or more nucleic acids. The term also includes any isolated nucleic acid sample, such as genomic DNA, fresh frozen or formalin-fixed paraffin-embedded nucleic acid specimens. It is also envisioned that the sample may come from: a single individual, a collection of nucleic acid samples from genetically related members, a nucleic acid sample from a genetically unrelated member, a nucleic acid sample (matched therewith) from a single individual (such as a tumor sample and a normal tissue sample), or a sample from a single source containing two different forms of genetic material (such as maternal DNA and fetal DNA obtained from a maternal subject), or there is contaminating bacterial DNA in a sample containing plant or animal DNA. In some embodiments, the source of nucleic acid material may include nucleic acids obtained from newborns, such as nucleic acids commonly used for newborn screening.

[0251] Nucleic acid samples may include high molecular weight substances, such as genomic DNA (gDNA). Samples may include low molecular weight substances, such as nucleic acid molecules obtained from FFPE samples or archived DNA samples. In another embodiment, low molecular weight substances include DNA of enzymatic fragmentation or mechanical fragmentation. Samples may include cell-free circulating DNA. In some embodiments, samples may include nucleic acid molecules obtained from biopsy tissue, tumor, scrapings, swabs, blood, mucus, urine, blood plasma, semen, hair, laser capture microdissection, surgical resection and other clinical or laboratory samples. In some embodiments, samples may be epidemiological samples, agricultural samples, forensic samples or pathogenic samples. In some embodiments, samples may include nucleic acid molecules obtained from animals (such as humans or mammalian sources). In another embodiment, samples may include nucleic acid molecules obtained from non-mammalian sources (such as plants, bacteria, viruses or fungi). In some embodiments, the source of nucleic acid molecules may be archived or extinct samples or species.

[0252] In addition, the methods and compositions disclosed herein can be used to amplify nucleic acid samples with low-quality nucleic acid molecules, such as degraded and / or fragmented genomic DNA from forensic samples. In one embodiment, forensic samples may include nucleic acids obtained from crime scenes, nucleic acids obtained from missing persons DNA databases, nucleic acids obtained from laboratories associated with forensic investigations, or include forensic samples obtained by law enforcement agencies, one or more military services, or any such personnel. Nucleic acid samples can be purified samples or lysates containing crude DNA, such as from oral swabs, paper, fabrics, or other substrates that can be impregnated with saliva, blood, or other body fluids. Therefore, in some embodiments, the nucleic acid sample may include a small amount of DNA (such as genomic DNA), or a fragmented portion of DNA. In some embodiments, the target sequence may be present in one or more body fluids, wherein body fluids include but are not limited to blood, sputum, plasma, semen, urine, and serum. In some embodiments, the target sequence may be obtained from the victim's hair, skin, tissue sample, autopsy, or remains. In some embodiments, nucleic acids comprising one or more target sequences may be obtained from dead animals or humans. In some embodiments, the target sequence may include a nucleic acid obtained from non-human DNA (such as microbial, plant, or insect DNA). In some embodiments, the target sequence or amplified target sequence is directed to the purpose of human identification. In some embodiments, the disclosure as a whole relates to methods for identifying the characteristics of forensic samples. In some embodiments, the disclosure as a whole relates to human identification methods using one or more target-specific primers disclosed herein or one or more target-specific primers designed using the primer design standards outlined herein. In one embodiment, a forensic sample or human identification sample containing at least one target sequence can be amplified using any one or more target-specific primers disclosed herein or using the primer standards outlined herein.

[0253] Components of the sequencing system 112 or the accelerated genotype imputation system 106 may include software, hardware, or both. For example, the sequencing system 112 or the accelerated genotype imputation system 106 may include one or more instructions stored on a computer-readable storage medium and executable by a processor of one or more computing devices (e.g., client devices 116). When executed by one or more processors, the computer-executable instructions of the sequencing system 112 or the accelerated genotype imputation system 106 may cause the computing device to perform the bubble detection method described herein. Alternatively, the sequencing system 112 or the accelerated genotype imputation system 106 may include hardware, such as a dedicated processing device that performs a specific function or a set of functions. Additionally or alternatively, the sequencing system 112 or the accelerated genotype imputation system 106 may include a combination of computer-executable instructions and hardware.

[0254] In addition, the components of the accelerated genotype imputation system 106 that perform the functions described herein for the accelerated genotype imputation system 106 may be implemented, for example, as part of a stand-alone application, as a module of an application, as a plug-in for an application, as one or more library functions that can be detected by other applications, and / or as a cloud computing model. Therefore, the components of the accelerated genotype imputation system 106 may be implemented as part of a stand-alone application on a personal computing device or a mobile device. Additionally or alternatively, the components of the accelerated genotype imputation system 106 may be implemented in any application that provides sequencing services, including but not limited to Illumina BaseSpace, Illumina DRAGEN, or Illumina TruSight software. "Illumina", "BaseSpace", "DRAGEN", and "TruSight" are registered trademarks or trademarks of Illumina, Inc. in the United States and / or other countries.

[0255] As discussed in more detail below, embodiments of the present disclosure may include or utilize a special-purpose or general-purpose computer including computer hardware (such as, for example, one or more processors and system memory). Embodiments within the scope of the present disclosure also include physical and other computer-readable media for carrying or storing computer-executable instructions and / or data structures. Specifically, one or more of the processes described herein may be implemented at least in part as instructions embodied in a non-transitory computer-readable medium and executable by one or more computing devices (e.g., any of the media content access devices described herein). In general, a processor (e.g., a microprocessor) receives instructions from a non-transitory computer-readable medium (e.g., a memory, etc.) and executes those instructions, thereby performing one or more processes, including one or more of the processes described herein.

[0256] Computer readable media can be any available media that can be accessed by a general or special purpose computer system. A computer readable medium that stores computer executable instructions is a non-transitory computer readable storage medium (device). A computer readable medium that carries computer executable instructions is a transmission medium. Therefore, by way of example and not limitation, embodiments of the present disclosure may include at least two distinct types of computer readable media: a non-transitory computer readable storage medium (device) and a transmission medium.

[0257] Non-transitory computer-readable storage media (devices) include RAM, ROM, EEPROM, CD-ROM, solid-state drives (SSD) (e.g., RAM-based), flash memory, phase-change memory (PCM), other types of memory, other optical disk storage devices, magnetic disk storage devices or other magnetic storage devices, or any other media that can be used to store desired program code means in the form of computer-executable instructions or data structures and which can be accessed by a general or special purpose computer.

[0258] "Network" is defined as one or more data links that enable the transmission of electronic data between computer systems and / or modules and / or other electronic devices. When information is transferred or provided to a computer via a network or another communication connection (hardwired, wireless, or a combination of hardwired or wireless), the computer appropriately regards the connection as a transmission medium. The transmission medium may include a network and / or data link, which can be used to carry the desired program code means in the form of computer executable instructions or data structures, and which can be accessed by general or special computers. The above combinations should also be included in the scope of computer-readable media.

[0259] Furthermore, upon reaching various computer system components, program code means in the form of computer executable instructions or data structures may be automatically transferred from a transmission medium to a non-transitory computer readable storage medium (device) (or vice versa). For example, a computer executable instruction or data structure received over a network or data link may be buffered in RAM within a network interface module (e.g., NIC), and then ultimately transferred to the computer system RAM and / or to a less volatile computer storage medium (device) at the computer system. Thus, it should be understood that a non-transitory computer readable storage medium (device) may be included in a computer system component that also (or even primarily) utilizes a transmission medium.

[0260] Computer executable instructions include, for example, instructions and data that make a general-purpose computer, a special-purpose computer, or a special-purpose processing device perform a certain function or function group when executed at a processor. In some embodiments, computer executable instructions are executed on a general-purpose computer to turn a general-purpose computer into a special-purpose computer that implements an element of the present disclosure. Computer executable instructions can be, for example, binary numbers, intermediate format instructions such as assembly language, or even source code. Although the subject matter has been described in a language specific to structural features and / or method actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the described features or actions. On the contrary, the described features and actions are disclosed as exemplary forms of implementing the claims.

[0261] Those skilled in the art will appreciate that the present disclosure can be practiced in a network computing environment with many types of computer system configurations, including personal computers, desktop computers, portable computers, message processors, handheld devices, multiprocessor systems, microprocessor-based or programmable consumer electronics, network PCs, minicomputers, mainframe computers, mobile phones, PDAs, tablet computers, pagers, routers, switches, etc. The present disclosure can also be practiced in a distributed system environment, where both local and remote computer systems linked by a network (by a hardwired data link, a wireless data link, or a combination of hardwired and wireless data links) perform tasks. In a distributed system environment, program modules can be located in both local and remote memory storage devices.

[0262] Embodiments of the present disclosure may also be implemented in a cloud computing environment. In this specification, "cloud computing" is defined as a model for implementing on-demand network access to a shared pool of configurable computing resources. For example, cloud computing may be employed in the marketplace to provide ubiquitous and convenient on-demand access to a shared pool of configurable computing resources. A shared pool of configurable computing resources may be quickly provisioned via virtualization and released with low management effort or service provider interaction, and then scaled accordingly.

[0263] The cloud computing model may consist of various characteristics, such as, for example, on-demand self-service, broad network access, resource pooling, rapid elasticity, metered services, etc. The cloud computing model may also exhibit various service models, such as, for example, software as a service (SaaS), platform as a service (PaaS), and infrastructure as a service (IaaS). The cloud computing model may also be deployed using different deployment models, such as private cloud, community cloud, public cloud, hybrid cloud, etc. In this specification and in the claims, a "cloud computing environment" is an environment in which cloud computing is employed.

[0264] Fig.13 1300 is a block diagram of a computing device 1300 that can be configured to perform one or more of the above processes. It will be appreciated that one or more computing devices such as computing device 1300 can implement the accelerated genotype estimation system 106. Fig.13 As shown, computing device 1300 may include a processor 1302, memory 1304, storage device 1306, I / O interface 1308, and communication interface 1310 that may be communicatively coupled via communication infrastructure 1312. In some embodiments, computing device 1300 may include Fig.13 Those shown may have fewer or more components. The following paragraphs describe in more detail Fig.13 Components of computing device 1300 are shown.

[0265] In one or more embodiments, the processor 1302 includes hardware for executing instructions such as those that constitute a computer program. As an example and not by way of limitation, in order to execute instructions for dynamically modifying a workflow, the processor 1302 may retrieve (or fetch) instructions from an internal register, an internal cache, a memory 1304, or a storage device 1306, and decode and execute them. The memory 1304 may be a volatile or non-volatile memory for storing data, metadata, and programs executed by the processor. The storage device 1306 includes a storage device for storing data or instructions for executing the methods described herein, such as a hard disk, a flash drive, or other digital storage device.

[0266] I / O interface 1308 allows a user to provide input to computing device 1300, receive output from the computing device and otherwise transfer data to the computing device and receive data from the computing device. I / O interface 1308 may include a combination of a mouse, a keypad or keyboard, a touch screen, a camera, an optical scanner, a network interface, a modem, other known I / O devices or such I / O interfaces. I / O interface 1308 may include one or more devices for presenting output to a user, including but not limited to a graphics engine, a display (e.g., a display screen), one or more output drivers (e.g., a display driver), one or more audio speakers and one or more audio drivers. In certain embodiments, I / O interface 1308 is configured to provide graphics data to the display for presentation to the user. Graphics data may represent one or more graphical user interfaces and / or may serve any other graphics content of a specific implementation.

[0267] The communication interface 1310 may include hardware, software, or both. In any case, the communication interface 1310 may provide one or more interfaces for communication (such as, for example, packet-based communication) between the computing device 1300 and one or more other computing devices or networks. As an example and not by way of limitation, the communication interface 1310 may include a network interface controller (NIC) or network adapter for communicating with an Ethernet or other wired-based network, or a wireless NIC (WNIC) or wireless adapter for communicating with a wireless network such as WI-FI.

[0268] Additionally, the communication interface 1310 can facilitate communication with various types of wired or wireless networks. The communication interface 1310 can also facilitate communication using various communication protocols. The communication infrastructure 1312 can also include hardware, software, or both that couple the components of the computing device 1300 to each other. For example, the communication interface 1310 can use one or more networks and / or protocols to enable multiple computing devices connected through a particular infrastructure to communicate with each other to perform one or more aspects of the processes described herein. For illustration, the sequencing process can allow multiple devices (e.g., client devices, sequencing devices, and server devices) to exchange information such as sequencing data and error notifications.

[0269] In the foregoing description, the present disclosure has been described with reference to specific exemplary embodiments thereof. Various embodiments and aspects of the present disclosure are described with reference to the details discussed herein, and the accompanying drawings illustrate various embodiments. The above description and drawings are illustrative of the present disclosure and should not be construed as limiting the present disclosure. Many specific details are described to provide a thorough understanding of the various embodiments of the present disclosure.

[0270] The present disclosure can be embodied in other specific forms without departing from its spirit or essential characteristics. The embodiments described should be considered as being only exemplary and not restrictive in all respects. For example, the methods described herein can be performed with fewer or more steps / actions, or the steps / actions can be performed in different orders. Additionally, the steps / actions described herein can be repeated or performed in parallel with each other or in parallel with different examples of the same or similar steps / actions. Therefore, the scope of the application is indicated by the appended claims rather than the foregoing description. All changes within the equivalent meaning and scope of the claims will be included within their scope.

Claims

1. A method, comprising: Using genotype imputation models to identify haplotype reference groups for genomic regions of genomic samples; accessing from a memory device and for a marker variant, a first transition-aware allele likelihood factor corresponding to a haplotype allele from the haplotype reference set and a second transition-aware allele likelihood factor corresponding to the haplotype allele; combining, by a configurable processor, the first transition-aware allele likelihood factor and a neighboring marker intermediate allele likelihood for the genomic region including the haplotype allele given a neighboring marker variant to generate a neighboring marker transition factor-aware allele likelihood for the marker variant and a haplotype from the haplotype reference set; determining, by the configurable processor and for the marker variant and the haplotype, an intermediate allele likelihood for the genomic region including the haplotype allele based on the neighboring marker transition factor-aware allele likelihood and the second transition-aware allele likelihood factor; as well as An allelic likelihood for the genomic region including haplotype alleles from the haplotype reference set is generated by the configurable processor and for the set of marker variants corresponding to the genomic region based on the intermediate allelic likelihood.

2. The method of claim 1, further comprising pre-determining the first transition-aware allele likelihood factor and the second transition-aware allele likelihood factor before determining one or more intermediate allele likelihoods corresponding to the marker variant.

3. The method according to claim 2, wherein: Predetermining the first transition-aware allele likelihood factor comprises combining the allele likelihood factor of the haplotype allele and a transition constant coefficient for transitioning between haplotypes from the haplotype reference set; and Predetermining the second transition-aware allele likelihood factor comprises combining the allele likelihood factor and a transition linear coefficient for transitioning between haplotypes from the haplotype reference set.

4. The method of claim 1 , wherein combining the first transition-aware allele likelihood factor and the neighboring marker intermediate allele likelihood comprises multiplying the first transition-aware allele likelihood factor and the neighboring marker intermediate allele likelihood without further multiplication to determine the intermediate allele likelihood.

5. The method according to claim 1, further comprising: accessing the second transition-aware allele likelihood factor as part of the summed neighboring marker transition-aware allele likelihood factors; as well as The intermediate allele likelihood is determined based on the neighboring marker transition factor-aware allele likelihood and the summed neighboring marker transition-aware allele likelihood factor.

6. The method according to claim 5 further includes predetermining the summed neighboring marker transition-aware allele probability factor by combining the allele probability factor of the haplotype allele, the transition constant coefficient for transitioning between haplotypes from the haplotype reference group, and the summed neighboring marker intermediate allele probability of the neighboring marker variants.

7. The method of claim 6, wherein the allele likelihood factor of the haplotype allele comprises a reference allele likelihood factor of a sample reference haplotype allele or an alternative allele likelihood factor of a sample alternative haplotype allele.

8. The method of claim 1, wherein the genotype estimation model comprises a Hidden Markov Genotype Estimation Model.

9. The method of claim 1, wherein the configurable processor comprises an application specific integrated circuit (ASIC), an application specific standard product (ASSP), a coarse-grained reconfigurable array (CGRA), or a field programmable gate array (FPGA).

10. The method of claim 1, wherein the memory device comprises a dynamic random access memory (DRAM), a dynamic random access memory (SRAM), or a cache memory device.

11. A system, comprising: at least one processor; Memory devices; as well as a non-transitory computer readable medium comprising instructions that, when executed by the at least one processor, cause the system to: Using genotype imputation models to identify haplotype reference groups for genomic regions of genomic samples; accessing from the memory device and for a marker variant, a first allele likelihood factor corresponding to a haplotype allele from the haplotype reference set and a second allele likelihood factor corresponding to the haplotype allele; combining said first allele likelihood factor and a neighboring marker intermediate allele likelihood for said genomic region including said haplotype allele given a neighboring marker variant to generate a neighboring marker factor-aware allele likelihood for said marker variant and a haplotype from said haplotype reference set; determining an intermediate allele likelihood for the genomic region including the haplotype allele for the marker variant and the haplotype based on the neighboring marker factor-perceived allele likelihood and the second allele likelihood factor; as well as An allelic likelihood for the genomic region including haplotype alleles from the haplotype reference set is generated based on the intermediate allelic likelihood for the set of marker variants corresponding to the genomic region.

12. The system of claim 11, further comprising instructions which, when executed by the at least one processor, cause the system to access the first allele likelihood factor and the second allele likelihood factor corresponding to the haplotype allele from the haplotype reference group and the second transition-aware allele likelihood factor corresponding to the haplotype allele from the memory device and for the marker variant by accessing the first allele likelihood factor and the second allele likelihood factor from the memory device and for the marker variant.

13. The system of claim 12, further comprising instructions that, when executed by the at least one processor, cause the system to predetermine the first transition-aware allele likelihood factor and the second transition-aware allele likelihood factor prior to determining one or more intermediate allele likelihoods corresponding to the marker variant as part of a round across a haplotype matrix.

14. The system of claim 13, further comprising instructions that, when executed by the at least one processor, cause the system to: predetermining the first transition-aware allele likelihood factor by combining the allele likelihood factor of the haplotype allele and a transition constant coefficient for transitioning between haplotypes from the haplotype reference set; and The second transition-aware allele likelihood factor is predetermined by combining the allele likelihood factor and a transition linear coefficient for transitioning between haplotypes from the haplotype reference set.

15. The system of claim 12, further comprising instructions that, when executed by the at least one processor, cause the system to determine the first transition-aware allele likelihood factor by combining an allele likelihood factor and a transition linear coefficient.

16. The system of claim 15, wherein: The first allele likelihood factor comprises an allele likelihood factor of a sample reference haplotype allele or a sample alternative haplotype allele; and The second allele likelihood factor includes the allele likelihood factor of the sample reference haplotype allele or the sample alternative haplotype allele.

17. The system of claim 11, further comprising instructions that, when executed by the at least one processor, cause the system to determine the intermediate allele probability by multiplying the first transition-aware allele probability factor and the adjacent marker intermediate allele probability to combine the first allele probability factor and the adjacent marker intermediate allele probability without further multiplication operations.

18. The system of claim 11, further comprising a data flow engine and instructions that, when executed by the at least one processor, cause the system to: Sending a corresponding set of input values ​​including allele likelihood factors, transition coefficients, and haplotype allele values ​​from the data flow engine to a corresponding accelerated computing engine in the accelerated computing engine cluster; and A respective set of intermediate allele possibilities corresponding to a respective subset of marker variants and a respective subset of haplotypes is determined by the respective accelerated computation engine and based on the respective set of input values.

19. The system of claim 18, further comprising instructions that, when executed by the at least one processor, cause the system to: The corresponding input value set is sent from the data flow engine to the corresponding accelerated computing engine by the following operations: Sending a first input value set including an allele likelihood factor, a transition coefficient, and a haplotype allele value from the data flow engine to a first accelerated computing engine in the accelerated computing engine cluster; Sending a second input value set including an allele likelihood factor, a transition coefficient, and a haplotype allele value from the data flow engine to a second accelerated computing engine in the accelerated computing engine cluster; as well as The corresponding intermediate allele possibility set is determined by the following operation: determining, by the first accelerated computing engine and based on the first set of input values, a first intermediate allele likelihood set corresponding to a first subset of marker variants and a first subset of haplotypes; as well as A second intermediate allele likelihood set corresponding to a second subset of marker variants and a second subset of haplotypes is determined by the second accelerated computing engine and based on the second set of input values.

20. The system of claim 11, further comprising instructions which, when executed by the at least one processor, cause the system to determine the intermediate allele likelihood by determining the intermediate allele likelihood for the genomic region that includes a sample reference haplotype allele or a sample replacement haplotype allele.

21. The system of claim 11, further comprising instructions which, when executed by the at least one processor, cause the system to determine the intermediate allele probability based on the neighboring marker factor perceived allele probability and the second allele probability factor by summing the neighboring marker transition factor perceived allele probability and the summed neighboring marker transition perceived allele probability factor.

22. The system of claim 11, further comprising instructions that, when executed by the at least one processor, cause the system to determine one or more nucleobase calls for the genomic region from the genomic sample based on the allele likelihood of the genomic region and one or more variant nucleobase calls surrounding the genomic region.

23. A method comprising: determining, with a configurable processor that performs the first round, a first round intermediate allele likelihood for a genomic region from a genomic sample comprising haplotype alleles corresponding to the set of haplotypes given a set of marker variants; storing a subset of first round intermediate allele likelihoods corresponding to a subset of marker variants of a marker variant set; regenerating, with the configurable processor, the first round intermediate allele likelihood by initializing allele likelihood determinations at the marker variant set using a subset of the stored first round intermediate allele likelihoods; determining, with the configurable processor performing a second round, a second round intermediate allele likelihood for the genomic regions comprising the haplotype alleles corresponding to the set of haplotypes given the set of marker variants; and An allele likelihood for the genomic region including the haplotype allele is generated based on the regenerated first round intermediate allele likelihood and the second round intermediate allele likelihood.

24. The method of claim 23, wherein the configurable processor comprises an application specific integrated circuit (ASIC), an application specific standard product (ASSP), a coarse-grained reconfigurable array (CGRA), or a field programmable gate array (FPGA).

25. The method of claim 23, wherein: Determining the first round intermediate allele likelihood includes determining a reverse intermediate allele likelihood for the genomic region including the haplotype allele using a reverse round; and Determining the second round intermediate allele likelihood includes using a forward round to determine a forward intermediate allele likelihood for the genomic region including the haplotype allele.

26. The method of claim 23, wherein: Storing the subset of first-round intermediate allele probabilities includes storing the subset of first-round intermediate allele probabilities in a dynamic random access memory (DRAM); and Initializing the allele likelihood determination at the marker variant set using the stored subset of first-round intermediate allele likelihoods includes accessing the stored subset of first-round intermediate allele likelihoods from the DRAM.

27. The method of claim 23, wherein generating the allele likelihood based on the regenerated first round intermediate allele likelihood and the second round intermediate allele likelihood comprises: determining a summed first-round intermediate allele likelihood for the set of marker variants based on the regenerated first-round intermediate allele likelihood; determining a summed second-round intermediate allele likelihood for the set of marker variants based on the second-round intermediate allele likelihood; as well as The allele likelihood is determined based on the summed first round intermediate allele likelihood and the summed second round intermediate allele likelihood.

28. The method according to claim 23, further comprising: determining a running sum of a first subset of intermediate allele likelihoods for said genomic region including haplotype alleles of a first type of one or more haplotypes from a haplotype reference set for neighboring marker variants; determining a running sum of a second subset of intermediate allele possibilities for the genomic region including haplotype alleles of a second type from the one or more haplotypes for the neighboring marker variants; as well as A sum of intermediate allele probabilities for the genomic region comprising haplotype alleles of haplotypes from the haplotype reference set is determined for marker variants based on the running sum of the first subset of intermediate allele probabilities and the running sum of the second subset of intermediate allele probabilities.

29. The method of claim 23, further comprising: storing haplotype allele indicator data of the haplotype matrix on a dynamic random access memory (DRAM); as well as The haplotype allele indicator data of the haplotype matrix is ​​accessed from the DRAM by the configurable processor to generate the allele likelihood using a hidden Markov haploid genotype imputation model or a hidden Markov diploid genotype imputation model.

30. The method of claim 23, wherein initializing allele likelihood determination at the marker variant set using a stored subset of first-round intermediate allele likelihoods comprises: determining a first subset of first round intermediate allele probabilities for the first marker variant set based on a first stored column of first round intermediate allele probabilities for initial marker variants from the first marker variant set; as well as A second subset of first round median allele probabilities for the second marker variant set is determined based on a second stored column of first round median allele probabilities for initial marker variants from the second marker variant set.

31. A system, comprising: at least one processor; Memory devices; as well as a non-transitory computer readable medium comprising instructions that, when executed by the at least one processor, cause the system to: by performing a first round of determining a first round intermediate allele likelihood for a genomic region from a genomic sample including haplotype alleles corresponding to the set of haplotypes given a set of marker variants; storing on the memory device a subset of first round intermediate allele likelihoods corresponding to a subset of marker variants of a set of marker variants; regenerating the first round intermediate allele likelihood by initializing allele likelihood determinations at the marker variant set using a subset of the stored first round intermediate allele likelihoods; by performing a second round of determining a second round intermediate allele likelihood for the genomic region comprising the haplotype allele corresponding to the set of haplotypes given the set of marker variants; as well as An output engine is used to generate an allele likelihood for the genomic region including the haplotype allele based on the regenerated first round intermediate allele likelihood and the second round intermediate allele likelihood.

32. The system of claim 31 , further comprising: a haplotype allele indicator memory, the haplotype allele indicator memory being used to store haplotype allele indicator data; A conversion coefficient memory, wherein the conversion coefficient memory is used to store the conversion coefficient; as well as An allele likelihood factor memory is used to store the allele likelihood factors.

33. The system of claim 31 further comprising a joint engine for determining intermediate allele likelihood values.

34. The system of claim 31 , further comprising a data flow engine and instructions that, when executed by the at least one processor, cause the system to: Sending a corresponding set of input values ​​including allele likelihood factors, transition coefficients, and haplotype allele values ​​from the data flow engine to a corresponding accelerated computing engine in the accelerated computing engine cluster; and A respective set of intermediate allele possibilities corresponding to a respective subset of marker variants and a respective subset of haplotypes is determined by the respective accelerated computation engine and based on the respective set of input values.

35. The system of claim 34, further comprising instructions that, when executed by the at least one processor, cause the system to: The corresponding input value set is sent from the data flow engine to the corresponding accelerated computing engine by the following operations: Sending a first input value set including an allele likelihood factor, a transition coefficient, and a haplotype allele value from the data flow engine to a first accelerated computing engine in the accelerated computing engine cluster; Sending a second input value set including an allele likelihood factor, a transition coefficient, and a haplotype allele value from the data flow engine to a second accelerated computing engine in the accelerated computing engine cluster; as well as The corresponding intermediate allele possibility set is determined by the following operation: determining, by the first accelerated computing engine and based on the first set of input values, a first intermediate allele likelihood set corresponding to a first subset of marker variants and a first subset of haplotypes; as well as A second intermediate allele likelihood set corresponding to a second subset of marker variants and a second subset of haplotypes is determined by the second accelerated computing engine and based on the second set of input values.

36. The system of claim 31, further comprising a data flow engine corresponding to a cluster of accelerated computing engines and instructions that, when executed by the at least one processor, cause the system to: Sending the subset of the first-round intermediate allele probabilities from the data flow engine to a first accelerated computing engine in the accelerated computing engine cluster so that the first accelerated computing engine can regenerate the first-round intermediate allele probabilities; and An additional subset of the first-round intermediate allele probabilities is sent from the data flow engine to a second accelerated computing engine in the accelerated computing engine cluster so that the second accelerated computing engine can regenerate additional first-round intermediate allele probabilities.

37. The system of claim 31 , further comprising a data flow engine and instructions that, when executed by the at least one processor, cause the system to: sending the subset of first round intermediate allele possibilities from the memory device to the data flow engine; and The subset of first-round intermediate allele probabilities is sent from the data flow engine to the accelerated computing engine to regenerate the first-round intermediate allele probabilities based on the subset of first-round intermediate allele probabilities.

38. The system of claim 31 , further comprising a data flow engine and instructions that, when executed by the at least one processor, cause the system to: storing haplotype allele indicator data of a haplotype matrix on the memory device; and The haplotype allele indicator data of the haplotype matrix is ​​accessed from the memory device to generate the allele likelihood using a hidden Markov haploid genotype imputation model or a hidden Markov diploid genotype imputation model.

39. The system of claim 31, wherein the memory device comprises a dynamic random access memory (DRAM), a dynamic random access memory (SRAM), or a cache memory device.

40. The system of claim 31 , further comprising instructions that, when executed by the at least one processor, cause the system to determine one or more nucleobase calls for the genomic region from the genomic sample based on the allele likelihood for the genomic region and one or more variant nucleobase calls surrounding the genomic region.

41. A method comprising: Using genotype imputation models to identify haplotype reference groups for genomic regions of genomic samples; determining a running sum of a first subset of intermediate allele likelihoods for said genomic regions including haplotype alleles of a first type of one or more haplotypes from said haplotype reference set for neighboring marker variants; determining a running sum of a second subset of intermediate allele possibilities for the genomic region including haplotype alleles of a second type from the one or more haplotypes for the neighboring marker variants; determining a sum of intermediate allele probabilities for said genomic region comprising haplotype alleles of haplotypes from said haplotype reference set based on said running sum of said first subset of intermediate allele probabilities and said running sum of said second subset of intermediate allele probabilities for marker variants; and An allele likelihood for the genomic region including the haplotype allele is generated based on the sum of the intermediate allele likelihoods.

42. The method of claim 41, wherein the first type of haplotype allele comprises a sample reference haplotype allele and the second type of haplotype allele comprises a sample replacement haplotype allele.

43. A method according to claim 41, wherein determining the sum of intermediate allele probabilities includes determining an initial intermediate allele probability by a configurable processor and based on the intermediate allele probability from the first subset of intermediate allele probabilities or from the second subset of intermediate allele probabilities for the marker variant and from the intermediate allele probabilities before summing the neighboring marker intermediate allele probabilities of the genomic region including the haplotype allele for the neighboring marker variant.

44. A method according to claim 41, wherein determining the sum of intermediate allele probabilities includes determining, by a configurable processor and for the marker variant, an initial intermediate allele probability based on the intermediate allele probability from the first subset of intermediate allele probabilities or from the second subset of intermediate allele probabilities and from the intermediate allele probabilities before generating the allele probability for the genomic region including the haplotype allele for the neighboring marker variant.

45. The method of claim 41, further comprising: predetermining a first transition-aware allele likelihood factor for a row corresponding to the first type of haplotype alleles and a second transition-aware allele likelihood factor for a row corresponding to the second type of haplotype alleles; as well as A sum of intermediate allele probabilities is further determined based on the first transition-aware allele likelihood factor for the row corresponding to the first type of haplotype allele and the second transition-aware allele likelihood factor for the row corresponding to the second type of haplotype allele.

46. ​​The method of claim 45, further comprising: determining, for the neighboring marker variants, a sum of neighboring markers for an intermediate allele likelihood for the genomic region including the haplotype allele; as well as The sum of intermediate allele probabilities is further determined for the neighboring marker variants based on a combination of the neighboring marker sum of intermediate allele probabilities, the first transition-aware allele probability factor for the row corresponding to the first type of haplotype allele, and the second transition-aware allele probability factor for the row corresponding to the second type of haplotype allele.

47. The method of claim 41, further comprising: multiplying the running sum of the first subset of intermediate allele likelihoods by a first transition-aware allele likelihood factor; multiplying the running sum of the second subset of intermediate allele likelihoods by a second transition-aware allele likelihood factor; as well as The sum of intermediate allele probabilities is determined for the marker variants based on the multiplied running sum of the first subset of intermediate allele probabilities and the multiplied running sum of the second subset of intermediate allele probabilities.

48. The method of claim 47, further comprising: Predetermining the first transition-aware allele likelihood factor comprises combining a first allele likelihood factor for a haplotype allele of the first type and a transition linear coefficient for transitioning between haplotypes from the haplotype reference set; and Predetermining the second transition-aware allele likelihood factor includes combining a second allele likelihood factor for haplotype alleles of the second type and the transition linear coefficient.

49. A non-transitory computer-readable medium storing instructions that, when executed by at least one processor, cause a computing device to: Using genotype imputation models to identify haplotype reference groups for genomic regions of genomic samples; determining a running sum of a first subset of intermediate allele likelihoods for said genomic regions including haplotype alleles of a first type of one or more haplotypes from said haplotype reference set for neighboring marker variants; determining a running sum of a second subset of intermediate allele possibilities for the genomic region including haplotype alleles of a second type from the one or more haplotypes for the neighboring marker variants; determining a sum of intermediate allele probabilities for said genomic region comprising haplotype alleles of haplotypes from said haplotype reference set based on said running sum of said first subset of intermediate allele probabilities and said running sum of said second subset of intermediate allele probabilities for marker variants; and An allele likelihood for the genomic region including the haplotype allele is generated based on the sum of the intermediate allele likelihoods.

50. The non-transitory computer readable medium of claim 49, wherein the first type of haplotype allele comprises a sample reference haplotype allele and the second type of haplotype allele comprises a sample replacement haplotype allele.

51. The non-transitory computer-readable medium of claim 49, further comprising instructions which, when executed by the at least one processor, cause the computing device to determine the sum of intermediate allele probabilities by: determining, by a configurable processor and for the marker variant, an initial intermediate allele probability based on intermediate neighboring allele probabilities from the first subset of intermediate allele probabilities or from the second subset of intermediate allele probabilities and from the intermediate allele probabilities before summing the intermediate allele probabilities for the genomic region including the haplotype allele for the neighboring marker variant.

52. The non-transitory computer-readable medium of claim 49, further comprising instructions which, when executed by the at least one processor, cause the computing device to determine the sum of intermediate allele probabilities by: determining, by a configurable processor and for the marker variant, an initial intermediate allele probability based on an intermediate neighboring allele probability from the first subset of intermediate allele probabilities or from the second subset of intermediate allele probabilities and from the intermediate allele probabilities before generating an allele probability for the genomic region including the haplotype allele for the neighboring marker variant.

53. The non-transitory computer readable medium of claim 49, further comprising instructions that, when executed by the at least one processor, cause the computing device to: predetermining a first transition-aware allele likelihood factor for a row corresponding to the first type of haplotype alleles and a second transition-aware allele likelihood factor for a row corresponding to the second type of haplotype alleles; and The sum of intermediate allele likelihoods is further determined based on the first transition-aware allele likelihood factor for the row corresponding to the first type of haplotype allele and the second transition-aware allele likelihood factor for the row corresponding to the second type of haplotype allele.

54. The non-transitory computer readable medium of claim 53, further comprising instructions that, when executed by the at least one processor, cause the computing device to: determining a sum of neighboring markers for the intermediate allele likelihood for the genomic region including the haplotype allele for the neighboring marker variants; and The sum of intermediate allele probabilities is further determined for the marker variant based on a combination of the neighboring marker sum of intermediate allele probabilities, the first transition-aware allele probability factor for the row corresponding to the first type of haplotype allele, and the second transition-aware allele probability factor for the row corresponding to the second type of haplotype allele.

55. The non-transitory computer readable medium of claim 49, further comprising instructions that, when executed by the at least one processor, cause the computing device to: multiplying the running sum of the first subset of intermediate allele likelihoods by a first transition-aware allele likelihood factor; multiplying the running sum of the second subset of intermediate allele likelihoods by a second transition-aware allele likelihood factor; and The sum of intermediate allele probabilities is determined for the marker variants based on the multiplied running sum of the first subset of intermediate allele probabilities and the multiplied running sum of the second subset of intermediate allele probabilities.

56. The non-transitory computer readable medium of claim 55, further comprising instructions that, when executed by the at least one processor, cause the computing device to: Predetermining the first transition-aware allele likelihood factor comprises combining a first allele likelihood factor for a haplotype allele of the first type and a transition linear coefficient for transitioning between haplotypes from the haplotype reference set; and Predetermining the second transition-aware allele likelihood factor includes combining a second allele likelihood factor for haplotype alleles of the second type and the transition linear coefficient.

57. The non-transitory computer-readable medium of claim 56, wherein the first allele likelihood factor comprises an allele likelihood factor for a sample reference haplotype allele, and the second allele likelihood factor comprises an allele likelihood factor for a sample alternative haplotype allele.

Images