Structural variant analysis

Abstract

The present disclosure provides methods, systems and algorithms for identifying and reporting genome or chromosome level structural information such as the presence of structural variation. In some cases, structural variation includes copy number variation, inversion, deletion, tandem duplication or inverted duplication. Further provided herein are methods, systems and algorithms for assembling paired read genome data, including creating and optimizing scaffold models.

Description

[0001] Cross-references

[0002] This application claims the benefit of U.S. Provisional Application No. 62 / 583,974, filed on November 9, 2017, which is hereby expressly incorporated by reference in its entirety. Background Art

[0003] It is still difficult to generate high-quality, highly continuous genome sequences in theory and practice. The problem becomes complicated when attempting to recover genome sequences, phasing information, or other required genetic information from samples preserved such as formalin-fixed, paraffin-embedded (FFPE) samples. Although the reduction in sequencing costs and time has increased the amount of available raw genome data, the lack of suitable methods to analyze and assemble data in an effective and accurate manner is a major limitation of current sequencing technologies.

[0004] Incorporation by reference

[0005] All publications, patents and patent applications mentioned in this specification are incorporated herein by reference to the same extent as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated by reference. All publications, patents and patent applications mentioned in this specification are incorporated herein by reference in their entirety and any references cited therein. Summary of the invention

[0006] Provided herein are methods for nucleic acid structural variant detection. Some such methods include: a) mapping read pair information onto a reference nucleic acid scaffold; b) assigning read pair positions to a first bin such that the read pair midpoint falls within the first bin nucleic acid position range and the read pair interval falls within the first bin interval range; and c) estimating copy number variation based on the mappability value of the first bin. In some cases, the method further includes normalizing the copy number variation. Additionally, the method further includes displaying mappability by plotting the mapped read density of two samples against each other.

[0007] Methods for nucleic acid structural variant detection are provided herein. Some such methods include: a) mapping read pair information onto a reference nucleic acid scaffold; b) assigning read pair positions to a first bin such that the read pair midpoint falls within the first bin nucleic acid position range and the read pair interval falls within the first bin interval range; c) generating a two-dimensional image of the read pair information; wherein each pixel represents a bin; d) calculating a z score for at least one set of four pixels sharing a common corner in the image; wherein the z score is represented by a contrast between adjacent pixels; and e) identifying candidate hits when the z score exceeds a threshold. In some cases, the reference nucleic acid scaffold is a genome. Typically, each data set is obtained from a different paired end read direction. It is considered that the candidate hit is selected from one or more of a translocation, an inversion, a deletion, a duplication, and an interchromosomal structural variation.

[0008] Provided herein is a system for modeling a mixture of allele variation in a sample.Some such systems include: a set of weighted genome scaffold models, wherein each genome scaffold model includes a set of weighted chromosomes, wherein each chromosome is a linear graph of bins in a genome scaffold; and a module for calculating the log-likelihood ratio of at least two genome scaffold models to predict whether the read pairs sampled by the library will fall into bins. In some cases, the system herein further includes at least one feature detector module, wherein the at least one feature detector module proposes candidate modifications to the genome scaffold model. Typically, at least one feature detector module determines the bin boundaries of a sequence variant. Considering that a sequence variant is selected from one or more of translocation, inversion, deletion and duplication. Typically, the system further includes a module for generating an alternative model based on the input from at least one feature detector module.

[0009] Provided herein is a method for modeling allelic variation in a sample. Some such methods include: a) generating a set of weighted genome scaffold models, wherein each genome scaffold model includes a set of weighted chromosomes, wherein each chromosome is a linear graph of bins in a genome scaffold; b) calculating a score based on the ability of the model to describe the read pair sequencing information mapped to a reference sequence, wherein a higher score value indicates a more predictive model; and c) iteratively adding additional models to maximize the score value. Considering that the read pair sequencing information includes one or more of inversions, translocations, duplications, and deletions. In some cases, the method further includes a detection feature, wherein the detection feature includes joining or separating bins in the model to increase the score value. Typically, the sample is a cancer cell.

[0010] Methods for nucleic acid structural variant detection are provided herein. Some such methods include: a) mapping read pair information onto a predicted nucleic acid scaffold; b) assigning read pair positions to a first bin such that the read pair midpoint falls within the first bin nucleic acid position range and the read pair interval falls within the first bin interval range; c) generating a two-dimensional image of the read pair information; wherein each pixel represents a bin; and d) identifying at least one feature in the two-dimensional image corresponding to two sequence fragments connected by a common connecting sequence fragment. Typically, the method includes assembling the two sequence fragments connected by the common connecting sequence fragment in the correct order. Sometimes, the method includes discarding features corresponding to false positives.

[0011] The method provided herein includes: mapping the read pair sequence information onto the sequence support; and identifying the local variation of the density of the multiple read pair symbols so mapped. In some cases, the method includes assigning the local variation of the density to the corresponding structural arrangement feature. Typically, the method includes reconstructing the sequence support, thereby reducing the local variation of the density. Sometimes, mapping the read pair sequence information onto the sequence support includes positioning the symbol indicating the read pair so that the distance of the symbol from the axis representing the sequence support indicates the distance from the mapping position of the first read of the read pair on the sequence support to the mapping position of the second read of the read pair on the sequence support, and the position of the symbol relative to the axis representing the sequence support indicates the average value of the mapping position of the first read of the read pair and the mapping position of the second read of the read pair. Sometimes, reconstructing the sequence support includes reordering at least some contiguous groups of the sequence support. Alternatively or in combination, reconstructing the sequence support includes reorienting at least one contiguous group of the sequence support. Typically, reconstructing the sequence support includes introducing a break into at least one contiguous group of the sequence support. Sometimes, the method further includes introducing a sequence present at one edge of the break onto the second edge of the break. In some cases, reconstructing the sequence scaffold includes translocating a segment of the first contig into an internal region of the second contig. Sometimes, mapping the read pair sequence information onto the sequence scaffold includes assigning the read pair information to a plurality of bins. Typically, identifying local variations in density includes identifying a region of local low density with a symbol. Alternatively, identifying local variations in density includes identifying a region of local high density with a symbol. Sometimes, identifying local variations in density includes identifying a density at a first position and a density at a second position, wherein the density at the first position and the density at the second position are significantly different. In some cases, the first position and the second position are adjacent. Typically, the first position and the second position are equidistant from the sequence scaffold. Sometimes, identifying local variations in density includes obtaining an expected density at the first position and an observed density at the first position. Typically, the expected density at the first position is a density predicted by a density gradient that decreases monotonically as the distance from the axis representing the sequence scaffold increases. Optionally, a local density variation equal to a portion of an integer value of the ploidy of the sample indicates an event of sample ploidy complement in the ratio. In some cases, the scaffold represents a cancer cell genome. Alternatively or in combination, the scaffold represents a transgenic cell genome. Optionally, the scaffold represents a gene-edited genome. Typically, after remodeling, the scaffold has an N50 that is increased by at least 20%.

[0012] Provided herein are methods comprising obtaining a scaffold comprising sequence scaffold information. Some such methods include: obtaining paired read information; deploying paired read information so that at least some read pair information is depicted to indicate the position of each read in the read pair relative to the scaffold, and indicating the distance from one read to another read mapped on the scaffold; and identifying local variations in the density of the deployed paired read information. In some cases, the method includes assigning local variations in density to corresponding structural arrangement features. Sometimes, the method includes reconfiguring the scaffold to reduce local variations. Typically, obtaining a scaffold comprising sequence scaffold information includes sequencing a nucleic acid sample. Alternatively or in combination, obtaining a scaffold comprising sequence scaffold information includes receiving digital information representing a nucleic acid sample. Sometimes, the method includes obtaining a predicted density distribution for the deployed read pair information. Typically, identification includes identifying a significant difference between the predicted density distribution and the depicted read pair information density. Alternatively or in combination, identifying local variations includes identifying a density perturbation having a density peak at the vertex of a right angle. In some cases, the vertex of the right angle points to the axis representing the scaffold. Typically, obtaining paired end read information includes cross-linking unextracted nucleic acids. Sometimes, obtaining paired end read information includes cross-linking nucleic acids bound in chromatin. Typically, the chromatin is natural chromatin. Alternatively or in combination, obtaining paired-end read information includes binding a nucleic acid to a nucleic acid binding portion. In some cases, obtaining paired-end read information includes generating reconstructed chromatin. Typically, deploying paired read information includes assigning read pair information to multiple bins. Sometimes, reconstructing a sequence scaffold includes reordering at least some contigs of a sequence scaffold. Alternatively or in combination, reconstructing a sequence scaffold includes redirecting at least one contig of a sequence scaffold. Sometimes, reconstructing a sequence scaffold includes introducing a break into at least one contig of a sequence scaffold. Typically, the method includes introducing a sequence at one edge of the break onto a second edge of the break. Sometimes, reconstructing a sequence scaffold includes translocating a segment of a first contig into an internal region of a second contig. In some cases, the scaffold represents a cancer cell genome. Sometimes, the scaffold represents a transgenic cell genome. Alternatively or in combination, the scaffold represents a gene-edited genome. Typically, after reconstruction, the scaffold has an N50 that is increased by at least 20%. Sometimes, a local density variation that is a portion of an integer value equal to the ploidy of a sample indicates an event of sample ploidy complementation in that proportion.

[0013] Methods for identifying structural rearrangements in samples relative to sequence scaffolds are provided herein. Some such methods include: mapping read pair sequence information onto the sequence scaffold; identifying local density variations having a right angled side pointing to an axis corresponding to the sequence scaffold and having bilateral symmetry along a line bisecting the right angled side; and classifying the sample as having a simple translocation relative to the sequence scaffold, the sequence scaffold comprising a segment length at least as long as the longest farthest mapped read of the local density variation from the translocation site.

[0014] Provided herein are methods for identifying structural rearrangements in samples. Some such methods include: mapping read pair sequence information onto a sequence scaffold; identifying a local density variation having a right-angled side pointing to an axis corresponding to the sequence scaffold; identifying a subregion of the local density variation that breaks bilateral symmetry along a line bisecting the right-angled side; and classifying the sample as having a translocation relative to the sequence scaffold, the sequence scaffold comprising a segment of the sequence to which the symmetry-restoring read pair population is to be mapped.

[0015] Provided herein are methods for identifying structural rearrangements relative to sequence scaffolds in a sample. Some such methods include: mapping read pair sequence information onto the sequence scaffold; identifying a local density variation having a right-angled side pointing to an axis corresponding to the sequence scaffold; obtaining an expected read pair density distribution curve; and identifying a scaffold segment to which a read pair containing the local density variation is mapped; repositioning the scaffold segment so that the read pair containing the local density variation is mapped to a region indicated by the expected read pair density distribution curve to obtain a density of the local density variation.

[0016] Provided herein is a computer monitor configured to display the results of any of the methods described herein.

[0017] Provided herein are computer systems configured to perform the computing steps of any of the methods described herein.

[0018] Provided herein are visual representations of mapped read pair data described herein or generated using the methods described herein.

[0019] Methods for detecting nucleic acid structural variants are provided herein. Some such methods include: mapping read pair information onto a predicted nucleic acid scaffold; obtaining a structural variant hypothesis; calculating a likelihood parameter that the structural variant hypothesis is consistent with the read pair information; and if the likelihood parameter of the hypothesis is greater than the second likelihood parameter of the second hypothesis, the nucleic acid sample is classified as having a structural variant hypothesis, wherein mapping the read pair information onto the predicted nucleic acid scaffold includes assigning a read pair position to the read pair so that the read pair is assigned to its midpoint on the predicted nucleic acid scaffold on one axis; and assigning a value corresponding to its read pair interval on the second axis to the read pair. Sometimes, the read pair includes a first segment mapped to a first region of the nucleic acid molecule and a second segment mapped to a second region of the nucleic acid molecule, wherein the first segment and the second segment are not adjacent and share a common phase. Typically, if the read pair midpoint falls within the first bin nucleic acid position range and the read pair interval falls within the first bin interval range, the read pair position is assigned to the first bin. In some cases, the first bin nucleic acid position range is a regular spacing of the predicted nucleic acid scaffold. Alternatively or in combination, the first bin spacing range is the logarithmic spacing of the complete spacing range of the read pair information. Sometimes, the first bin nucleic acid range is a regular spacing of the nucleic acid scaffold, and wherein the first bin spacing range is the logarithmic spacing of the complete spacing range of the read pair information. In some cases, if the read pair midpoint falls within the second bin nucleic acid position range and the read pair spacing falls within the second bin spacing range, the read pair position is assigned to the second bin. Typically, substantially all read information is binned. Sometimes, calculating the likelihood parameter includes determining a likelihood contribution of the first bin. Typically, the likelihood contribution of the first bin includes a first likelihood factor proportional to the count of the read pairs mapped to the first bin. Alternatively or in combination, the likelihood contribution of the first bin includes a second likelihood factor proportional to the area of ​​the first bin. Sometimes, the likelihood contribution of the first bin includes a first likelihood factor proportional to the count of the read pairs mapped to the first bin, and wherein the likelihood contribution of the first bin includes a second likelihood factor proportional to the area of ​​the first bin. Typically, the method includes determining the likelihood contribution of a second bin that does not overlap with the first bin in area. Sometimes, the likelihood parameter includes the likelihood contribution of the first bin and the likelihood contribution of the second bin. Occasionally, the likelihood parameter includes the likelihood contribution of the third bin. Alternatively or in combination, the likelihood parameter includes the likelihood contribution of the reads to the information for substantially all bins. Sometimes, the hypothesis includes a structural variation having a left edge and a length. Typically, the structural variation has an orientation of at least one of a deletion, an inversion, a duplication in the same direction, an outward inverted repeat, and an inward inverted repeat. Occasionally, the second hypothesis includes a structural variant that is different in at least one of the left edge, the length, and the structural orientation. Sometimes, the nucleic acid structural variant is homozygous in the nucleic acid sample. Alternatively, the nucleic acid structural variant is heterozygous in the nucleic acid sample.

[0020] Methods for displaying putative structural variations in nucleic acid samples are provided herein. Some such methods include the steps of assigning a population of sequence reads to a population of numbered bins, and assigning a likelihood parameter for reads containing structural variation edges falling within a first bin of the population of bins, wherein the likelihood parameter for the first bin includes a first likelihood component and a second component, the first likelihood component including the number of reads mapped to the first bin, and the second component including the area of ​​the first bin. Sometimes, the method includes plotting the likelihood of the structural variation according to the number of bins. Typically, the likelihood parameter for the first bin includes a convolution of a first likelihood component and a second component, the first likelihood component including the number of reads mapped to the first bin, and the second component including the area of ​​the first bin. Alternatively or in combination, the likelihood parameter includes a likelihood component associating a structural variant prediction with the number of reads mapped to the first bin and a likelihood component including the area of ​​the first bin. Occasionally, the bin population shares a common bin width across a fixed nucleic acid distance. Sometimes, there is variation in bin height between members of the bin population. Typically, when plotted on a logarithmic axis, the bin heights appear constant. Often, the likelihood parameter relates to the probability that a sequence read comprising a junction having a structural variation with a left edge and a length maps to the first bin. Sometimes, the structural variation has an orientation of at least one of a deletion, an inversion, a direct repeat, an outward inverted repeat, and an internal inverted repeat. Typically, the sequence reads include read pairs. Occasionally, the read pair includes a first segment mapped to a first region of the nucleic acid molecule and a second segment mapped to a second region of the nucleic acid molecule, the first segment and the second segment being non-adjacent and sharing a common phase.

[0021] Provided herein is a method for identifying structural variants in a nucleic acid sample. Some such methods include the following steps: obtaining mapped read pair data of a nucleic acid sample; obtaining a nucleic acid scaffold sequence; obtaining likelihood probability information for each of a plurality of structural variant hypotheses comparing the read pair data to the nucleic acid scaffold sequence; and identifying the most likely hypothesis among the structural variant hypotheses; wherein the method evaluates at least 10Mb of nucleic acid scaffold sequence per minute. Often, the method includes mapping the read pair information onto the nucleic acid scaffold sequence; obtaining a structural variant hypothesis; calculating a likelihood parameter that the structural variant hypothesis is consistent with the read pair information; and if the likelihood parameter of the hypothesis is greater than the second likelihood parameter of the second hypothesis, the nucleic acid sample is classified as having a structural variant hypothesis. Occasionally, mapping the read pair information onto the nucleic acid scaffold sequence includes: assigning a read pair position to the read pair so that the read pair is assigned to its midpoint on the predicted nucleic acid scaffold on one axis; and the read pair is assigned a value corresponding to its read pair interval on the second axis. Typically, the read pair includes a first segment mapped to a first region of the nucleic acid molecule and a second segment mapped to a second region of the nucleic acid molecule, the first segment and the second segment being non-adjacent and sharing a common phase. Sometimes, if the read pair midpoint falls within the first bin nucleic acid position range and the read pair spacing falls within the first bin spacing range, the read pair position is assigned to the first bin. Occasionally, the first bin nucleic acid position range is a regular spacing of the nucleic acid scaffold. Typically, the first bin spacing range is a logarithmic spacing of the complete spacing range of the read pair information. Alternatively or in combination, the first bin nucleic acid position range is a regular spacing of the nucleic acid scaffold, and wherein the first bin spacing range is a logarithmic spacing of the complete spacing range of the read pair information. In some cases, if the read pair midpoint falls within the second bin nucleic acid position range and the read pair spacing falls within the second bin spacing range, the read pair position is assigned to the second bin. Often, substantially all of the read information is binned. Typically, calculating the likelihood parameter includes determining a likelihood contribution of the first bin. Occasionally, the likelihood contribution of the first bin includes a first likelihood factor proportional to the count of the read pairs mapped to the first bin. Sometimes, the likelihood contribution of the first bin includes a second likelihood factor proportional to the area of ​​the first bin. Alternatively or in combination, the likelihood contribution of a first bin comprises a first likelihood factor proportional to the count of read pairs mapped to the first bin, and wherein the likelihood contribution of the first bin comprises a second likelihood factor proportional to the area of ​​the first bin. Often, the method further comprises determining a likelihood contribution of a second bin that does not overlap in area with the first bin. Typically, the likelihood parameters comprise the likelihood contribution of the first bin and the likelihood contribution of the second bin. Sometimes, the likelihood parameters comprise the likelihood contribution of a third bin. Occasionally, the likelihood parameters comprise likelihood contributions of read pair information for substantially all bins. Typically, the hypothesis comprises structural variation having a left edge and a length.In some embodiments, the second hypothesis comprises at least one structural variant that is different from the first one in the left edge, the length and the structural orientation. Occasionally, the nucleic acid structure variant is isozygous in the nucleic acid sample. Alternatively, the nucleic acid structure variant is heterozygous in the nucleic acid sample.

[0022] Methods for selecting a treatment regimen are provided herein. Some such methods include: performing the method of any of the foregoing embodiments, identifying the rearrangement, and identifying a treatment regimen consistent with the rearrangement. Often, the treatment regimen includes drug administration. Alternatively or in combination, the treatment regimen includes tissue resection.

[0023] Provided herein are methods for evaluating a treatment regimen. Some such methods include: performing the method of any of the foregoing embodiments a first time, administering the treatment regimen, and performing the treatment regimen a second time. Occasionally, the method includes suspending the treatment regimen. Alternatively, the method includes increasing the dosage of the treatment regimen. Sometimes, the method includes reducing the dosage of the treatment regimen. Alternatively, the method includes continuing the treatment regimen. Often, the treatment regimen includes medication. Typically, the treatment regimen includes surgical intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The patent or application file contains at least one drawing printed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee.

[0025] Figure 1 An exemplary schematic diagram of a protocol for analyzing read pair library data is depicted.

[0026] Figure 2A , Figure 2B and Figure 2C Depicted is a visual representation of the read pair library data used for copy number variant estimation.

[0027] Figure 2D Depicted is a visual representation of the copy number variation between the two samples.

[0028] Figure 3A Depicted is a visual representation of the mapped read pair data for samples matched to scaffolds as a plot of read pair spacing relative to the midpoint position of the mapped read pairs.

[0029] Figure 3B Depicted is a visual representation of mapped read pair data for samples with inversions as a plot of read pair spacing relative to the midpoint position of the mapped read pairs.

[0030] Figure 3CDepicted is an expanded scale visual representation of mapped read pair data for a sample with an inversion as a plot of read pair spacing relative to the midpoint position of the mapped read pairs.

[0031] Figure 3D Depicted is a graphical representation of the mapped read pair end data for the heterozygous inversion between points a and b.

[0032] Figure 4A Illustrations depicting various types of structural variation and the types of mapped read pair density patterns that result.

[0033] Figure 4B Depicted is a general illustration of the mapped read pair data observed for structural variation.

[0034] Figure 4C A general illustration of the mapped read pair data for the deletions observed is depicted.

[0035] Figure 4D Depicted is a general illustration of the mapped read pair data observed for inversions.

[0036] Figure 4E Depicted is a general illustration of the mapped read pair data observed for direct tandem repeats.

[0037] Figure 4F A general illustration of the mapped read pair data observed for the inverted tandem repeat R is depicted.

[0038] Figure 4G A general illustration of the mapped read pair data observed for the inverted tandem repeat L is depicted.

[0039] Figure 5A Depicted is a visual representation of the mapped read pair data as a plot of log likelihood ratio versus number of bins for a data set containing an inversion.

[0040] Figure 5B Depicted is a visual representation of the mapped read pair data as a plot of log likelihood ratio versus number of bins for a data set with regions of LLR approximately 0.

[0041] Figure 5C Depicted is a visual representation of the mapped read pair data as a plot of log-likelihood ratio versus number of bins for a dataset with regions without structural variation.

[0042] Fig. 6A and Figure 6B An exemplary simple kernel that can be used to find reciprocal translocations is depicted.

[0043] Figure 6CDepicted is a method for analyzing features using the ratio of foreground (fg) and background (bg) regions.

[0044] Fig.6D Images with features identified using the z-score method are depicted.

[0045] Figure 7 Depicted is an image of read pair data mapped onto scaffolds illustrating intra-chromosomal rearrangements.

[0046] Fig. 8A Describes the “secondary connection (2 nd Figure 2. Illustration of a "degree link" assembly where two different assembly outcomes are possible by analyzing only the first-degree link read pairs.

[0047] Figure 8B , Figure 8C and Fig.8D Illustration depicting the "Secondary Connection" assembly situation using feature detection.

[0048] Fig. 8E Two plots are depicted showing the contribution of abundance of read pairs in the mixture (γ) and gap size / distance (g) in the variation of predicted mapped read pair density (contours).

[0049] Fig. 9 Images with features corresponding to the reciprocal translocation between ETV6 and NTRK3 are depicted.

[0050] Fig. 10A , Fig. 10B and Fig. 10C Depicted are the results of image-based analysis of the same chromosome pair compared in three different samples.

[0051] Fig.11A , Fig. 11B and Fig. 11C Depicts chromosome 1 vs chromosome 7 ( Fig.11A ), chromosome 2 pair, chromosome 5 ( Fig. 11B ) and chromosome 1 to chromosome 1 ( Fig. 11C ) of the normalized read density (over 10 samples).

[0052] Fig. 12A and Fig. 12B Various bin handling methods are depicted. Fig. 12A shows equal bin sizes, while Fig. 12B Bin cell interpolation is shown.

[0053] Fig.13 Analysis via the whole genome scanning analysis pipeline is depicted.

[0054] Fig.14A and Fig. 14B Depicted are read pairs from the FFPE-based 'Chicago' library ( Fig.14A ) and the classic ‘Chicago’ based read pair library ( Fig. 14B ) to read the distance frequency data.

[0055] Fig.15A and Fig. 15B The mapping positions of read pairs mapped near structural differences between GM12878 and the reference on the GRCh38 reference sequence are illustrated. Fig.15A Data for an 80 kb inversion with flanking 20 kb duplication regions are depicted. Fig. 15B Data for phased heterozygous deletions are depicted.

[0056] Fig.16A Depicted are the shifted segment deviations in the mapped read pair data compared to a reference scaffold. In this case, the vertical segment of the data (vertical line) has been shifted to an alternative "hole" portion of the graph (arrow).

[0057] Fig. 16B Depicted are the folded segment biases in the mapped read pair data compared to the reference scaffold. In this case, segments B and B' both map to the same adjacent segment A on the scaffold.

[0058] Fig. 16C Depicted are collapsed duplications and misjoining biases in the mapped read pair data compared to a reference scaffold. In this case, highly similar sequences B / X are collapsed into a single assembly in the scaffold.

[0059] Fig.17A Depicted is an exemplary workflow for iteratively refining a genome scaffold model to improve the quality of read pair data mapped on the scaffolds.

[0060] Fig. 17B An image depicting read pair data mapped onto scaffolds prior to model optimization for potato chromosomes.

[0061] Fig. 17C An image depicting read pair data mapped onto scaffolds after model optimization for potato chromosomes.

[0062] Fig.18A An exemplary computer system is shown that is programmed or otherwise configured to implement the methods provided herein.

[0063] Fig.18B An example of a computer system that may be used in conjunction with exemplary embodiments of the present invention is shown.

[0064] Fig.18Cis a block diagram illustrating a first exemplary architecture of a computer system 700 that may be used in conjunction with exemplary embodiments of the present invention.

[0065] Fig.18D is a schematic diagram illustrating a network 2100 configured to incorporate multiple computer systems, multiple cell phones and personal data assistants, and network attached storage (NAS) that may be used in conjunction with exemplary embodiments of the present invention.

[0066] Fig.18E is a block diagram of a multiprocessor computer system 900 that employs a shared virtual address memory space that may be used in conjunction with exemplary embodiments of the present invention. DETAILED DESCRIPTION

[0067] Disclosed herein are methods and systems related to detecting, visualizing, and correcting rearrangements relative to sequence scaffolds, as indicated by analysis of nucleic acid samples. In some cases, the rearrangements indicate molecular events that occur in some or all samples when evaluated in comparison to a human reference genome, such as genomic rearrangements that frequently occur in human or other cancer cells. Alternative "rearrangements" relevant to the present disclosure include draft or even previously published genome assemblies for which a large amount of contig information is available, but for which one or more contigs may be mispositioned, e.g., placed out of order, misoriented relative to an experimentally determined sample, have highly similar folded regions, or are constructed using incorrectly connected contig components.

[0068] In both cases, practice of the methods and systems herein allows identification of differences (if any) between pre- or concurrently generated sequence information scaffolds and data indicating short-range and long-range physical linkage information experimentally generated by generating read pairs. The differences described herein are generally referred to as kernels, features, or symbols.

[0069] Phase information, chromosome conformation, sequence assembly and genetic characteristics, including but not limited to structural variation (SV), copy number variation (CNV), loss of heterozygosity (LOH), single nucleotide variant (SNV), single nucleotide polymorphism (SNP), chromosome translocation, gene fusion and insertion and deletion (INDEL), can be determined by analyzing the sequence read data generated by the method disclosed herein. Other inputs for analyzing genetic characteristics can include a reference genome (e.g., with annotations), genome masking information, and a list of candidate genes, gene pairs and / or coordinates of interest. Configuration parameters and genome masking information can be customized, or default parameters and genome masking can be used.

[0070] The methods described herein employ various steps associated with the processing of sequencing data. Optionally, each step utilizes the results or considerations from the previous step and produces a result or output. In some cases, steps in the method workflow are omitted or replaced by other steps. In some cases, sequencing data (such as data generated according to Hi-C or other paired reading schemes) are obtained by processing and sequencing samples. Exemplary steps for analyzing sequencing data generally include read mapping (mapping paired sequence reads from an individual to a reference), read binning (grouping reads by one or more properties), copy number estimation (copy number variation, CNV), normalization, de novo feature detection, breakpoint refinement, candidate scoring and reporting ( Figure 1 ). These steps are presented as examples only, as other steps for identifying and reporting features are also used with the methods and systems described herein.

[0071] Read pair generation

[0072] A variety of read pair generation methods are consistent with the disclosure herein. In an exemplary embodiment, read pairs are generated using "Hi-C" or related methods using natural or reconstructed chromatin to preserve the connection information between internally cut nucleic acid molecules so that the first region and the second region of the molecule are kept together regardless of their common phosphodiester backbone. However, the methods and systems herein are consistent with read pair data from a wide range of sources, and not all embodiments are limited to one or another read pair generation source.

[0073] Mapping read pair data

[0074] Common to many of the systems and methods herein is the generation of a binned read pair array, optionally presented as a two-dimensional map relative to a scaffold sequence axis. Local density variation on such a map is identified, and contigs of read pairs that cause local density variation are rearranged, reoriented, broken into fragments, or otherwise manipulated to reconstruct the scaffolds to which they contribute, thereby reducing overall or local density variation in the read pair binned array or read pair distribution map.

[0075] As used herein, a read pair data set is "mapped" to a sequence scaffold when the read pair data is binned or positioned relative to the scaffold sequence. In some cases, the mapped data is depicted spatially, such as depicted on a computer monitor, or printed out. Alternatively, the read pair data set mapped to the sequence scaffold is stored as a data array on a data storage medium of a computer. The read pair data is preferably "binned" or assigned to a specific position within a two-dimensional space or data array. Optionally, in a computer-generated image of the mapped read pair data set, the bins are represented by pixels.

[0076] The spatially mapped data is preferably presented so that the map positions of read pair intervals and individual reads of a read pair are captured in the positioning of symbols representing read pairs or occupied bins in the map.

[0077] For example, some methods of read pair data mapping include assigning read pairs to bins that are positioned such that the distance from the bin measured perpendicular to an axis representing the scaffold sequence corresponds to or indicates the separation between the locations where a first read and a second read of the read pair are most strongly mapped or aligned on the scaffold sequence. That is, read pairs with reads that align closely to each other on the scaffold are assigned to bins that are close to the axis, while read pairs with reads that are separated by a greater distance from each other are assigned to bins that are further away from the axis representing the sequence scaffold.

[0078] Optionally combined, read pairs are positioned along an axis representing a scaffold sequence so that they are assigned positions or bins having the closest point along the axis that approximately or exactly represents the midpoint between the scaffold position to which the first read is mapped and the scaffold position to which the second read is mapped. Depending on the representation of the data, the axis may be referred to as a central axis or a diagonal (axis). In some cases, the axis will be displayed horizontally, vertically, diagonally, or in any other configuration.

[0079] In the example of visualization, read pairs are mapped to genome scaffolds, and each pair is represented as a point in a plane whose x and y coordinates are equal to the distance between the matched read pairs. The xy plane can be divided into non-overlapping square bins, and the number of read pairs mapped to each bin can be tabulated. Bin counts can be displayed as an image (e.g., a heat map) of bins corresponding to pixels. In some cases, data from read pairs mapped as described herein are displayed as a 2D map with a horizontal axis or with an intensity corresponding to the read density. In some cases, data and / or identification features are processed without a visualization step.

[0080] A low degree of "background" is often observed in binning or read pair mapping. Such background manifests itself as a single "night sky" bin point in an empty portion of a data array or map visualization. Quantitatively, this background manifests itself as very low local bin density in areas of the map or data array that are expected or otherwise indicated to be devoid of read pairs.

[0081] Many technical factors separate from the disclosure herein explain such a "night sky" background. Factors include read pair sequence quality, "GC percentage" or base pair bias of the sample or scaffold, global or local repetitiveness of the genome, stringency of read-scaffold alignment, or other technical parameters.

[0082] Errors in read sequence base calling can cause reads to align to regions of the scaffold other than the region from which the underlying molecule was actually derived. Skewed GC percentages or repetitiveness can increase the chance that a read will align to multiple locations or that a single base error in sequencing can cause a read to align to the wrong region of the scaffold. These chances can be reduced by adjusting the stringency of base calling in sequencing or increasing the stringency of assigning reads to genomic regions.

[0083] However, increasing stringency at these steps or elsewhere in the sequence generation and alignment process may also exclude a large amount of accurate, informative data from the analysis. Therefore, in a given implementation of the methods or systems disclosed herein, individual samples, sequencing protocols, organisms, or experimental goals may determine the degree of tolerance for the "night sky" background.

[0084] Local density variation determination

[0085] According to the methods disclosed herein, it is often beneficial to evaluate local density variation in a read pair data array or a mapped read pair data set. Many methods can be used to evaluate local density variation to identify features, such as kernels in a data set array or a mapped data set.

[0086] The local density variation is assessed using many methods known to those skilled in the art. For example, the local density is determined and compared to the density of the immediately adjacent region of the mapped read pair data set or read pair array. Alternatively, the local density is compared to the density of a region positioned at a comparable or similar distance perpendicular to the axis defined by or corresponding to the scaffold sequence.

[0087] Instead of or in addition to a single comparison region, local density variation is optionally detected by comparing the local density to the average density along a line or band passing through the local region and parallel to the axis representing the scaffold sequence. That is, the local density is compared to the density of read pairs that share a common or comparable read pair interval but are distributed at other positions throughout the scaffold.

[0088] Alternatively or in combination, density values ​​are determined for each position in the entire map or data set such that the density is compared to the local density of at least one other position of the map or data set (such as 1, 2, 3, 4, 5 or more than 5 positions). The local density is determined and evaluated relative to the local density of at least one other position of the map or data set such that local density variation can be matched to positions on the map or data set with a common density, regardless of the distance from the axis or the average read pair distance of its members.

[0089] Similarly, in some cases, a density gradient is determined, such as a density gradient that decreases with distance from an axis (such as an axis representing a sequence scaffold). The local density is then compared to the density of the gradient, and if the local density is significantly different in density gradient value at a distance from the axis comparable to the distance of the local density region to the axis, the local density is classified as a "variant". "Significant" differences can be evaluated by any number of statistical, computational or other methods known in the art or consistent with the disclosure herein.

[0090] Following such a determination, in some cases, a "density predicted" position of the read pairs responsible for that local density is determined such that repositioning of a scaffold component (such as a contiguous group) on the axis causes the read pairs to be positioned so that the local density matches or is closer to the local density of the read pairs after repositioning of the scaffold or scaffold contiguous group.

[0091] Repositioning of contigs or other scaffold components is performed to reduce local density variation as assessed above, or to reduce an overall measure of density variation relative to an overall expected density gradient. Repositioning variously includes reordering scaffold components such as contigs relative to each other, reorienting at least one contig relative to a second contig, breaking a contig into at least two components, introducing sequences such as sequences adjacent to a break into a breakpoint boundary, or excising a segment (or fragment) from a contig sequence and introducing the segment to another location in the scaffold contig.

[0092] In some aspects, various modeling methods for predicting density are used to calculate the expected density variation. Optionally, a model for γ (mixture abundance) and g (gap size) is used, where the contours indicate the expected rate of change (or gradient) of density. In this model, the areas with the steepest density changes (contours) are often found to have low abundance / low gap size ( Fig. 8E , left) and high abundance / high gap size ( Fig. 8E , right). Additional models, including models based on empirically acquired data obtained from the methods and systems described herein, can also predict changes in density and are optionally incorporated throughout.

[0093] In some cases, local density is defined as being "close to" or "deviating from" a defined region on a mapped read pair map. In some cases, a region defined as "close to" a central axis corresponds to a region with an expected read density within at least 0.5X, 0.75X, 1X, 1.25X, 1.5X, 2X, or 2.5X of the average expected density just on the central axis. In some cases, a region defined as "deviating from" a central axis corresponds to a region with an expected read density that is no more than 0.1X, 0.2X, 0.3X, 0.4X, 0.5X, 0.75X, or no more than 0.9X of the average density on the central axis. Alternatively, a region defined as "close to" an axis is described in terms of the read pair spacing distance (in base pairs) from the central axis. Optionally, a read pair distance of at least 1, 2, 5, 10, 20, 50, 100, 200, 500, 1000, 2000, 5000, 10,000, 20,000, 50,000, 100,000, 200,000, 500,000, 1 million, 2 million, 5 million, 10 million, or at least 20 million base pairs from the central axis is defined as "off" the axis. In some cases, read pair distances of about 1, 2, 5, 10, 20, 50, 100, 200, 500, 1000, 2000, 5000, 10,000, 20,000, 50,000, 100,000, 200,000, 500,000, 1 million, 2 million, 5 million, 10 million, or about 20 million base pairs from the central axis are defined as "off" the axis. Similarly, read pair distances of no more than 1, 2, 5, 10, 20, 50, 100, 200, 500, 1000, 2000, 5000, 10,000, or no more than 20,000 base pairs from the central axis are defined as "close" to the axis. Similarly, read pair distances that are about 1, 2, 5, 10, 20, 50, 100, 200, 500, 1000, 2000, 5000, 10,000, or about 20,000 base pairs from the central axis are defined as being "close" to the axis. Alternatively, read pair distances are represented by bins, where each bin represents a range of read pair distances in base pairs.

[0094] In various manifestations of the methods described herein, the read density between two defined regions is compared to establish the boundaries or presence of a kernel. In some cases, the difference is at least 10%, 20%, 50%, 80%, 100%, 200%, 500%, 800%, 1000%, 2000%, 5000%, or at least 5000%. In other cases, the difference is about 10%, 20%, 50%, 80%, 100%, 200%, 500%, 800%, 1000%, 2000%, 5000%, or at least 5000%.

[0095] In various manifestations of the methods described herein, the read density difference ("higher" or "lower") between the observed density and the expected density is compared to identify the deviation between the model scaffold and the mapped read pair data. In some cases, the difference is at least 10%, 20%, 50%, 80%, 100%, 200%, 500%, 800%, 1000%, 2000%, 5000% or at least 5000%. In other cases, the difference is about 10%, 20%, 50%, 80%, 100%, 200%, 500%, 800%, 1000%, 2000%, 5000% or at least 5000%.

[0096] Complex rearrangement assessment

[0097] In some cases, analysis of the reads on the bin array or map indicates that the bin distribution is consistent with a particular rearrangement relative to the sequence scaffold. Often, a particular rearrangement has multiple effects or features on the bin array or map, depending on the extent or correlation of multiple events of the rearrangement on a molecule (such as a chromosome) or in a predicted sequence (such as a scaffold sequence).

[0098] After identifying local density variations in a data array or map indicating a rearrangement, details of investigating secondary local density variations or local density variations indicating the extent or co-occurrence of multiple events in a rearrangement are taught by some methods and systems herein. For example, a simple translocation event results in a characteristic local density distribution, and a symmetrical local density distribution will be produced if the fragment length involved in the occurrence is greater than the density resolution of the map or binned data array. However, if the translocation or scaffold rearrangement is an internal segment rather than a complete arm of a molecule or scaffold, one or more perturbations may be seen assuming that the segment is within the density resolution of the map or binned data array. The local density distribution indicating the event may lack bilateral symmetry along a line that bisects the local density variation at the point closest to the axis. Alternatively or in combination, a second local density distribution is detected, which involves a read pair having a read mapped to a region where the read is expected, which region, if mapped to the first local density variation, will restore symmetry with the previous local density variation. Such a density distribution is often indicative of a complex rearrangement in the sample molecule or scaffold such that the two breakpoints join three different segments relative to the starting or intended scaffold.

[0099] exist Fig. 8A An exemplary complex rearrangement "secondary connection" situation is shown in FIG. Fig. 8A , top) separate at the indicated sites to form fragments (labeled ag) and rearrange to form products ( Fig. 8A, bottom). The analysis was complicated by the co-ligation of fragments A and G with fragment D, which would have produced signals consistent with ade / cdg and adg recombination fragments. However, in some cases, the analysis was performed by identifying Figure 8B exists in Fig. 8A The additional long-range signal ag of adg that is not present in (ade / cdg) is used to distinguish these two cases. In some cases, other methods are used to reduce the possibility of false positive fusion decisions caused by observing these long-range signals ( Fig.8D In one approach to reducing false positives, all fusion calls are grouped by shared breakpoints and rejected if they share two breakpoints with a higher-scoring call. In another approach to reducing false positives, a model-based discriminant approach is applied to examine the likelihood based on γ (mixture abundance) and g (gap size). Fig. 8E ), where the contours predict the expected rate of change of density.

[0100] Local density variation geometry

[0101] Local density variation typically manifests itself in the mapping output as having at least one right-angled edge "pointing" toward the axis, so that the line that locally bisects that angle represents the shortest distance from the local density variation to the axis.

[0102] Some local density variations are square, showing bilateral symmetry along lines drawn perpendicular to the axis and bisecting the right-angled side pointing toward the axis.

[0103] Alternatively, some local density variations exhibit bilateral symmetry as described above, but have a difficult to define distal edge or boundary because the local density variation is significantly larger at right angles to the directional axis relative to elsewhere in the local density variation.

[0104] Alternatively, some local density variations are rectangular rather than square, lacking bilateral symmetry along a line drawn perpendicular to the axis and bisecting the right angle side pointing toward the axis. In extreme cases, such local density variations appear linear at lower levels of resolution. Additionally, local density variations with configurations other than those described above have been observed.

[0105] Alternatively, some local density variations are "bow-tie" shaped, with a center point defined approximately midway between segment lengths and equidistant from the axis. In some cases, four density regions are observed that intersect at right angles at the center point, with boundary lines of these regions intersecting the axis at 45 degree angles and crossing segment boundaries on the axis. One region of density is optionally bounded by the axis, and in some cases, regions adjacent to a region bounded by the axis have higher than expected density.

[0106] Information from local density

[0107] The methods and systems disclosed herein allow for the use of local density determination towards multiple ends in the various methods herein.

[0108] The peak change of local density variation, such as that seen at the right angles closest to the axis representing the scaffold sequence, provides information in some cases as a measure of the copy number of the genomic event associated therewith. That is, the local density variation indicating adjacent segments, either alone or in combination with other atlas or bin array information, is analyzed with respect to the peak density. The density is compared to the peak density immediately outside the axis of the atlas or data set. The metrics used variously include mean, median, mode, or other measures of density on the axis.

[0109] In some cases, the comparison of the ratio of an integer to another integer indicates the ploidy of the event related to the local density variation. That is, the density of half of the local axis density indicates the event of monoploid in the diploid sample. The density of one eighth of the local axis density indicates the event that occurs on a chromosome of an octoploid sample. The density of five eighths of the local axis density indicates the event that occurs on five chromosomes of an octoploid sample. Other combinations are obvious to those skilled in the art, such as 1 / 4, 2 / 4 or 3 / 4 in a tetraploid genome, 1 / 8, 2 / 8, 3 / 8, 4 / 8, 5 / 8, 6 / 8, 7 / 8 or 8 / 8 in an octoploid genome, 1 / 6, 2 / 6, 3 / 6, 4 / 6, 5 / 6 or 6 / 6 in a hexaploid genome, or other ratios related to or approximate integer ratios in a range consistent with the sample genome ploidy. Similarly, in some cases, the heterogeneity of the gene set also can cause the integer variation of local density. For example, a density that occurs at 1 / 10 of the expected density for a haploid sample indicates that 1 / 10 of the genome includes the event. These events are often present in heterogeneous cell populations, such as tumors or other diverse cell populations.

[0110] Alternatively or in combination, in some cases, the peak density of the local density variation seen at the right angles of the axis closest to the representative scaffold sequence is used as a measure of the distance between the edges of the genomic events involved in the scaffold sequence to provide information. That is, the local density variation of the indicated physical connection segment, either alone or in combination with other atlases or box element array information, is analyzed with respect to the peak density. The density is compared with the density gradient within the background density range that starts from the immediately adjacent axis outside the atlas or data set and decreases to the further away from the axis. The metric used includes the mean, median, mode or other on-axis density measurements to determine the points on the density gradient.

[0111] The density of local density variation is determined and compared to the read pair bin density gradient to find the off-axis distance on the gradient with comparable density. The scaffold sequence is then reconfigured to position the reap pairs of local density variation so that their density matches that of the gradient. Thus, the scaffold components are reconfigured to reduce the overall density variation in the data array or atlas relative to the gradient.

[0112] For a set of ideal read pairs data mapped on an ideal support, almost all density is evenly distributed on the central axis. Alternatively, a data model is used to predict the distribution of density, thereby generating an expected density or density gradient reduced from an axis. In some cases, a high-density or low-density region indication relative to the expected density on the diagonal axis indicates a difference between the read pair data and the support model. For example, in some cases, there is a folded fragment in the region indicating the support model with a density higher than expected on-axis. In another example, in some cases, there is a misconnection between two fragments in the region indicating the support model with a density lower than expected on-axis. On the one hand, misconnection connects two chromosomes together by mistake. In some respects, the density variation on the axis describes the deviation of any number between the read pair data and the support model observed.

[0113] Mathematical Model of Density

[0114] In one aspect of density data processing, genomic positions are plotted relative to read pair spacing (e.g., represented by the midpoint positions of mapped read pairs). In a genome without structural variation (SVs, differences, features, etc.), most points are distributed around the baseline ( Figure 3A ). However, the presence of variations such as inversions produces Figure 3B and Figure 3C Depicted plot. Regions lacking dots near the baseline indicate the edges of inversion segments. In some cases, structural variation is modeled as features or kernels, such as Figure 3D As shown, where sites a and b are the edges of the event, the light-colored points represent those points that are currently reflected above the midpoint of a and b (the intersection of the dashed lines), which are typically used to identify features. Optionally, a likelihood ratio is calculated by comparing the following hypotheses: 1) the SV is present in the genome, and 2) the genome matches the reference. In some cases, the hypothesis h is formulated as a linear operation, including representing the data in the region of interest as a set of read pair counts C in bins ij And A ij Set to the area of ​​each bin, thus calculating the log-likelihood ratio (LLR) contribution (S) of each read pair of bin i, j h ij ), and calculate the log-likelihood contribution per unit area of ​​bin i, j (T h ij). In one exemplary equation, the LLR score is expressed as:

[0115] S h =∑ ij S h ij C ij +∑ ij T h ij A ij

[0116] In some cases, it is beneficial to compute likelihood ratios for multiple SVs. h ij , T h ij ) Search for SVs at each offset k in the genome:

[0117] S h k =∑ i，j S h i，j C i-k，j +∑ i，j T h j，j A i-k，j

[0118] This process is optionally repeated to calculate likelihood ratios for all SVs in the genome.

[0119] In another example, analyzing Figure 4A As an example only, each variant including inversions, deletions, tandem duplications, and inverted repeats has a read pair mapping with an apparent spacing d0 and a possible true spacing d in the genome. i In some cases, Figure 4B-Figure 4G In the depicted variation, d is determined for each of the four regions (0, 1, 2, 3). i .

[0120] Read pair interval variations are often transformed into kernel elements using, for example, the Chicago likelihood model represented by the following equation:

[0121]

[0122] Where n represents the hits of the "rare" outcome in N attempts, and p is the total probability of the rare outcome:

[0123]

[0124] m is the multiplicity of alternative cases in the case of repetition.

[0125]

[0126] Or optionally, for the heterozygous case:

[0127]

[0128]

[0129] Occasionally, a bin will overlap with a feature or kernel region boundary. A possible solution involves computing the area and centroid of each overlapping region, using S h i,j max() and T h i,j As will be appreciated by those skilled in the art, alternative characteristic analysis equations and algorithms may also be used with the methods and systems herein.

[0130] Additional analysis techniques, such as image processing techniques, are variously used to identify signatures of genetic features, such as different rearrangements. For example, by analyzing a two-dimensional graph of paired reads, kernel convolution filtering can be used to find points in the image corresponding to fused pairs of genomic loci. Fig. 6A and Figure 6B An exemplary simple kernel that can be used to find mutual translocations is shown. In each case, the local z-score of the kernel is calculated by calculating the z-score contrast value, which is defined as the ratio of the foreground to background areas of the kernel, and repeating for each pixel ( Figure 6C ). Fig.6D Exemplary images with features identified from z-scores (circled) are shown. In some cases, reciprocal translocations between ETV6 and NTRK3 were identified ( Figure 7 ). The "bow tie" shaped features in the upper right and lower left quadrants indicate interactions between these two regions of the genome that have reciprocal translocation features. In some aspects, interchromosomal rearrangements are identified using a method of local z-score detection. Optionally, the process is repeated for each pixel in the image. In some cases, all local maxima that exceed a threshold are considered candidate hits for a feature.

[0131] Bracket Modeling

[0132] In some cases, the relationship between nucleic acid fragments (contigs, clusters, etc.) is represented by a mathematical graph model, in which each sequence is a node, and the interface between any two fragments in the assembly is represented as an edge connecting two or more nodes. In some cases, the path connecting all nodes by the edge (and only crossing each node once) represents the solution of the assembly of the sequencing fragment. Generally, the lack of unique overlapping regions when sequencing data fragments leads to multiple solutions (or paths) for the assembly. For example, in the ideal haploid series of fragments A, B and C, it is envisioned that 6 different options (or paths) connect all three fragments in a linear manner. However, if the edge between nodes A / B and B / C is shown as a kernel on the density map of the mapped read pairs on or near the central axis, and the scaffold model corresponds to the arrangement ABC, then the model accurately matches a single path ABC. In some cases, there is no density corresponding to the feature in the region corresponding to the edge (e.g., edge A / B), and the arrangement currently contains a "blocking edge" that informs the scaffold model, and reduces the number of possible paths. In some cases, the blocking edge prevents the path from being defined between two nodes of the graph model, thereby informing the assembly that the two fragments are not adjacent. Optionally, a weighting factor is given to each edge, which specifies the possibility of using the edge as part of the solution path. In some cases, the weighting factor represents the possibility of two node connections. For the scaffold model of ABC, in some cases, a lower density than expected will be observed on the diagonal of the expected AB feature, which will reduce the weighting factor of edge AB. In a practical sense, this allows in some cases to simplify the number of paths through nodes of the sequence graph model. In another example, a feature corresponding to edge AC is observed at the intersection of a horizontal line that bisects the position of segment A on the axis and a vertical line that bisects the position of segment C on the axis. For the scaffold model of ABC, this indicates in some cases that node (or segment) B is incorrectly placed in the scaffold model between segments A and C that should be adjacent.

[0133] Often, more complex translocation events are aided by the addition of blocking edges. Fig. 8A Two different rearrangements / pathways are depicted (left and right), each with an edge connecting fragments a / d and d / g. This assembly case and a variety of others are typically handled by applying graphical theoretical models. Blocking edges are added between a / g corresponding to the density of reads that lack mapping (top concentric circles, Figure 8B ), there is most likely only one path connecting ade and cdg. Alternatively, by adding blocking edges between a / e and c / g (two sets of concentric circles, Figure 8C), given the lack of density in the two regions represented by the concentric circles, only a single path corresponding to adg is possible. Optionally, more complex translocation events are also analyzed using this general strategy.

[0134] Model Evaluation

[0135] In some aspects, the method can be used to describe the entire support, chromosome or genome composed of many fragments (nodes), for which many assembly solutions represented by the path through the node are evaluated. Typically, variants exist as intrachromosomal variants and are processed using various data analysis methods (such as modeling defined by multiple potential equations). In an exemplary method of data analysis, a genome model "support" is constructed from a sequencing data set (such as a Hi-C data set). Optionally, data are obtained from a tumor, and include a mixture of a genome or any other sample of allele heterozygosity. In some aspects, a group of genomes comprising a high degree of genetic heterogeneity (such as a tumor) is modeled as a weighted genome model set, defined by the following equation:

[0136]

[0137] where each genome (G1, G2, etc.) is defined as a weighted (weighting factor α) model of a set of chromosomes. In some cases, each chromosome (C) is defined as a linear graph of bins on the genome:

[0138]

[0139] In some embodiments, the number of read pairs that map to a pair of genomic bins (i, j) is defined as a Poisson distribution:

[0140]

[0141] An example equation for the log-likelihood ratio of two models predicting reads for λ1 and λ2, respectively, is:

[0142]

[0143] In some aspects, the model provides the probability that a read pair sampled from a genome by a library falls into bin i, j. For the isotropic model (without transactivation domain (TAD)), the probability is optionally expressed as:

[0144]

[0145] where d g i,jis the shortest path distance between bins i and j in genome g, while p(d) is the empirical read path interval distribution. Alternatively or in combination, the read pair probability is specified with the copy number and mappability terms of bins i and j. In some cases, a non-isotropic model incorporating position-specific TADs is used:

[0146]

[0147] or in a more general form:

[0148]

[0149] Modifications and improvements to the model generally improve the quality and accuracy of the data. Typically, new components are added to the model to enhance its ability to describe the data. For example, a series of models Mk are generated to improve an initial model generated from a reference scaffold or a comparative genomic scaffold. It is usually assumed that a new genome is added to with weight γ, and the weights αi (1 < i < k) are updated to (1 - γ)αi respectively. Given multiple candidates for Mk+1, in some cases, the candidate that results in the largest increase in the score ΔS is selected:

[0150]

[0151] For example, in some cases, the best model is found by choosing γ that maximizes ΔS. Alternatively or in combination, all weights αi are adjusted to obtain an increased ΔS.

[0152] In some aspects, new candidates for mixture components are obtained that, when summed over all (i, j), result in larger ΔS values. However, the contributions of these potential model components to ΔS are typically concentrated in the ij plane near the fusion junction. In some cases, local image filtering identifies candidate edits. When such a local search identifies a high-scoring (and therefore not explained by the current model) contact between bins r and s, the contact is optionally added to a new "genome" or as an edit to a genome already present in the mixture. In some cases, feature detection methods propose candidate modifications to the model to explain the discovered features. For example, a basic set of feature detection methods includes one or more of the following: "reciprocal translocation +", "reciprocal translocation -", "translocation ++", "translocation + -", "translocation - +", "translocation - "or "break" methods. Feature detector methods typically output features, such as: break after bin i, break before bin j, or connect bin i to bin j. In some cases, the method obtains a list of features and models and generates alternative models for scoring. For example, if the model already consists of n alternative gene sets, the method optionally applies the edits of the feature to each of the n alternative gene sets, and makes a new copy of each alternative model to apply the edits for a total of 2n alternative models. Other scoring models may also be utilized during implementation of the method.

[0153] In another characterization technique, modeling is used to identify intrachromosomal rearrangements. For example, the likelihood of a rearrangement occurring is often determined by calculating the log-likelihood ratio (LLR), which is the ratio between two hypotheses:

[0154]

[0155] in is the expected number of reads in the 2D contact plane region under assumption i, is the probability of sampling a read pair with the spacing indicated by hypothesis i for read pair j, given the insert size distribution model. In some cases, the hypothesis is the signal of background mixed with background plus frequency λ. In some aspects, the assumptions are: a) the variant is present in the genomic region being analyzed, and b) the genome matches the reference. For example, an LLR score S is calculated for two hypotheses: (1) the reads are generated from a mixture of genomes where a fraction contain fusions between loci i and j relative to the reference, and (0) no such contact exists near i, j.

[0156]

[0157] The score contributed by n reads involving two bins separated by a gap d0 on the genome is typically expressed as follows (making a bin approximation), where the reads are positioned relative to the tested contact (i, j) such that the reads are separated by d1 in the rearranged genotype (a small region of the 2D contact plane):

[0158]

[0159] The score S is the in-plane sum of the contributions dS in w bins in each direction i, j.

[0160]

[0161] In some cases, a score "S" with respect to γ ​​estimates the variant abundance. In the limit of γ→1, this becomes separable and can be calculated using kernel convolution:

[0162]

[0163] where M is the matrix of observed read counts, K S1 is an element with ln P(d k,l ), K0 is a trivial kernel whose elements are equal to 1 and covers the kernel footprint, Q is the null hypothesis read likelihood contribution whose elements are equal to the element-wise product of M and P(d) (similar to the diagonal distance contour), N K 1 is a constant representing the number of reads expected from the rearranged genotypes within the kernel, and N0 is a matrix whose elements indicate the number of reads expected under the assumption of 0 (diagonal contours). For the first order of 1→γ,

[0164]

[0165] In some cases, this equation (e.g., γ<1) is approximated as

[0166]

[0167] is reasonable because Usually small, where P(d k,l )>>P(d i+k,j+l ).

[0168] In some aspects, the likelihood function determines the order and orientation of the contig. In some cases, the likelihood function is derived from the polynomial probability of observing a particular configuration of N balls falling into k+1 bins numbered 0, 1, ..., k, where x i is the number of balls (or paired-end reads) that fall into the i-th bin, and Pi is the probability that a ball will fall into the i-th bin:

[0169]

[0170] In one example, bin 0 has a higher probability than the remaining "rare" bins. If n << N balls fall into m "rare" bins and the remaining N - n balls end up in bin 0, the probability is typically described as

[0171]

[0172] where j indexes the rare bins that receive balls. Without loss of generality, in some cases, the bins are renumbered 1…k such that the first m of them are the bins that are hit by balls. P i xi The remaining factors of P (for bins where i > m and x i = 0) are all equal to 1. Optionally, further assume that the rare bins are so rare that none of them are hit by more than one ball and m = n, reducing the equation to:

[0173]

[0174] By the normalization condition on Pi, define p as the combined probability of all rare bins for convenience:

[0175]

[0176] According to the Poisson limit theorem, if N is very large and p is very small:

[0177]

[0178] where λ = Np. In some aspects, this simplifies the combinatorial factors in the probability expression. In some cases, make the substitution n = k and rewrite the approximation as:

[0179]

[0180]

[0181] In some cases, the log probability is expressed as:

[0182]

[0183] In some cases, P i is normalized to The Poisson approximation of the binomial distribution that controls n is typically used as long as N is large and and assuming that at most one ball falls into a given rare bin, which is typically reasonable. In some cases, the log-likelihood ratio is expressed as:

[0184]

[0185] In some cases, optimization of the scaffold model results in a decrease in the score S, indicating a model that better describes the data. The optimization process is optionally repeated until all deviations between the model and the mapped read pair data are eliminated. Fig.17A In, it is observed that an exemplary workflow for improving the scaffold model comprises the following steps: obtaining raw connection density data, generating contact probability scores, editing a side graph, generating distance fields, and updating contact probability relative to the current side graph. In some cases, the process results in an interactively updated genome model based on a graph. In some cases, the process is iterated to improve the quality of the mapping reads for feature identification to the data. In some cases, a contact probability score is generated for each potential feature (or deviation) in the graph. In some cases, the side graph editing refers to the weight given to the edge in the graph model of the change assembly, which affects the most likely assembly solution. In some aspects, these side graph editings correspond to reordering the fragments in the scaffold, removing fragments, repeating fragments or destroying fragments to establish better consistency between the scaffold model and the read pair data. Once edited, the shortest path through the graph model is usually identified, and the read pair data is mapped to the new scaffold model. In another step, the scaffold model is re-evaluated with all potential deviations between the read pair data, and a new score is generated. Optionally, these steps are repeated to minimize the overall score, indicating a more accurate scaffold assembly. In some cases, the overall effect is observed visually, for example, before optimizing the model Fig. 17B and later obtained Fig. 17C The difference between.

[0186] Other equations and methods for genomic modeling and expression probability are also used with the methods and systems described herein.

[0187] Copy number estimation

[0188] Calculation of copy number variation is often beneficial for assessing disease status, for example, in assessing gene copies with mutations associated with cancer. Copy number estimates for mutations are determined using a variety of methods, such as methods involving density assessment of local density variation relative to other regions or locations of a map or relative to a density gradient region. In some cases, copy number variation is calculated using the following equation:

[0189]

[0190]

[0191] Where N i is the number of mapped reads in bin i, N is the total number of mapped reads, w is the bin width, G is the genome size, c iis the number of copies of bin i, m i is the mappability of bin i. In some aspects, mappability refers to the ability to reassemble portions of a genome, and in some cases, highly repetitive sequences can hinder this ability. In some cases, if N i and m i are both smaller, then c i Biased towards 1. In some cases, chromosomes are divided into bins, and mapped read pairs are sorted into bins based on the midpoint of the pair. In some cases, the number of read pairs connecting genomic bins i and j follows the following equation:

[0192] N ij ~P(c i c j m i m j N p ij )

[0193] In some cases, 2D histograms were generated to visually display copy number data for different samples ( Figure 2A-2C ). On the other hand, the 2D histograms were normalized to separate the signals of long-range contacts from the following copy number differences:

[0194]

[0195] Often two or more samples are compared to show the effect of mappability. Figure 2D The sample CT407 is plotted relative to each other on each axis ( Figure 2A , left) and CT410( Figure 2A , right). In some aspects, the points falling outside the diagonal represent the copy number difference between the two samples compared. Alternatively or in combination, the above steps are performed without the aid of visualization and are alternatively stored on a non-transitory computer medium. Those skilled in the art will appreciate that alternative equations are also used to estimate copy number differences.

[0196] Sequencing

[0197] Input, such as sequence read data, can be formatted in an appropriate file format. For example, the sequence read data can be contained in a FASTA file, a FASTQ file, a BAM file, a SAM file, or other file format. The input sequence read data can be unaligned. The input sequence read data can be aligned.

[0198] The sequence read data can be prepared for analysis. For example, the reads can be trimmed for quality. If necessary, the reads can also be trimmed to remove sequencing adapters.

[0199] Sequence read data can be aligned. For example, read pairs can be aligned with a specified reference genome. In some cases, the reference genome is GRCh38. A variety of algorithms or tools can be used to perform alignment, including but not limited to SNAP, Burrows-Wheeler aligners (e.g., bwa-sw, bwa-mem, bwa-aln), Bowtie2, Novoalign, and modifications or variations thereof.

[0200] The quality control (QC) report of the analysis can also be generated.Before conducting a deeper sequencing, the QC report can be used to identify the failed library.Such quality control report can include multiple indicators.QC indicators can include but are not limited to total reads, repeats (for example, PCR repeats) percentage, unmapped read percentage, with low mapping quality (for example, Q<20) reading percentage, the percentage of readings mapped to different chromosomes, (such as the distance between the mapping positions) between 0 and 1kbp read to the percentage of insertion, between 1kbp and 100kbp read to the percentage of insertion, between 100kbp and 1Mbp read to the percentage of insertion, higher than 1Mbp read to the percentage of insertion, containing the reading of the connection joint right percentage, with the proximity of the restriction fragment end, read to the estimated value of the interval map and the complexity of the library.QC indicators can be used to optimize the analysis, and identify the quality problems of reagents, samples and users.Sequence alignment can be filtered based on one or more QC indicators.It is also possible to filter repeated readings, such as based on the comparison of the reading at the closely corresponding position.

[0201] The sequence read analysis result may include connection density results. The connection density results may include full genome, a locus and two locus views of the connection density results. The connection density results may be output as a data set. The connection density results may be presented as a connection density map (LDP), such as a heat map of the interaction (e.g., contact) between a chromosome or a genomic region. The connection density results may be associated with a score such as a quality score. In some cases, the connection density is output for a result exceeding a scoring threshold and is visualized. In an example, for the full genome, for judging from the beginning exceeding a scoring threshold, for judging the one-sided candidate exceeding a scoring threshold and for including all bilateral candidates that are classified as negative, their visualization is included. The connection density visualization may include a scale (e.g., a color scale), a length scale bar, a gene name label, the exon / intron structure symbol of a gene, and the highlighted display of the rearrangement detected.

[0202] The linkage information can be normalized to control for effects and biases such as coverage, fragment mappability, fragment GC content, and fragment length. Normalization can be performed by matrix balancing or other factor-agnostic methods. Matrix balancing can employ algorithms such as the Sinkhorn-Knopp algorithm or the Knight-Ruiz normalization. Normalization can also be performed to correct for background signals that may cause false positives. For example, Fig. 10A , Fig. 10B and Fig. 10C The results of image analysis based on the same chromosome pair compared in three different samples are shown. Several "hits" (circled in the figure) were found in the same position of multiple samples, which makes people suspect that these are false positives. Normalization, such as the normalization performed by the normalized read density median using a sample pool (e.g., 10 samples), can be used to correct single sample data, such as by dividing the sample pixel by the median pixel. Fig.11A , Fig. 11B and Fig. 11C Chromosome 1 to chromosome 7 ( Fig.11A ), chromosome 2 pair, chromosome 5 ( Fig. 11B ) and chromosome 1 to chromosome 1 ( Fig. 11C ) is the median normalized read density (on 10 samples). Normalization can be performed using various binning methods, including equal bin sizes, such as Fig. 12A As shown, and box element interpolation, such as Fig. 12B In some cases, bin interpolation can produce reduced background noise and result in higher resolution features compared to equal bin sizes.

[0203] The rearrangement of the sequence data of the comparison can be analyzed, including the rearrangement by the full genome and the rearrangement at the specific double locus (or bilateral) candidate gene. The analysis can also include identification of contact, fusion and engagement. The comparison of the sequence read data (for example, with BAM file or other suitable formats) can be input into the analysis. The genome masking information can also be input, or the genome masking information of default can be used in the analysis. It can be analyzed in the whole genome. In addition or alternatively, the list of bilateral candidate fusions can be analyzed. In some cases, the analysis performed on the candidate fusion list is more sensitive than the analysis performed on the full genome. The analysis of the bilateral candidate fusion can detect the fusion of the transposition involving the relatively short DNA segment that may be missed by the full genome scan.

[0204] In some cases, the distance measurement is performed as a combination of bases and base pairs. The minimum distance between breakpoints of a detectable rearrangement can be less than, about, or a number in a range defined by two numbers selected from a list of nucleic acid lengths including 2 bp, 3 bp, 4 bp, 5 bp, 6 bp, 7 bp, 8 bp, 9 bp, 10 bp, 20 bp, 30 bp, 40 bp, 50 bp, 60 bp, 70 bp, 80 bp, 90 bp, 100 bp, 200 bp, 300 bp, 400 bp, 500 bp, 600 bp, 700 bp, 800 bp, 900 bp, 1 kb, 2 kb, 3 kb, 4 kb, 5 kb, 6 kb, 7 kb, 8 kb, 9 kb, 10 kb, 20 kb, 30 kb, 40 kb, 50 kb, 60 kb, 70 kb, 80 kb, 90 kb, 20kb, 30kb, 40kb, 50kb, 60kb, 70kb, 80kb, 90kb, 100kb, 200kb, 300kb, 400kb, 500kb, 600kb, 700kb, 800kb, 900kb, 1Mb, 2Mb, 3Mb, 4Mb, 5Mb, 6M b, 7Mb, 8Mb, 9Mb, 10Mb, 20Mb, 30Mb, 40Mb, 50Mb, 60Mb, 70Mb, 80Mb, 90Mb , 100Mb, 200Mb, 300Mb, 400Mb, 500Mb, 600Mb, 700Mb, 800Mb, 900Mb or 1Gb.

[0205] The rearrangement analysis can generate a list of breakpoint pairs that are considered to be connected in the subject genome. The list of breakpoint coordinate pairs can also include a statistical significance or confidence measure (e.g., p-value) for the breakpoint coordinate pairs. These breakpoint pairs can be exported in an appropriate format, such as Browser Extensible Data (BED) or BED-PE.

[0206] The analysis of chromosome conformation can also be carried out using technology disclosed herein. For example, the domain (TAD) and TAD boundary of topological association can be determined. Other topological domains and boundaries can also be determined, including but not limited to the domain (LAD) associated with the layer, the replication time zone and the large organized chromatin K9-modification (LOCK) domain.

[0207] Fig.13 Analysis by the whole genome scanning analysis pipeline is shown. Sample calls made by the analysis pipeline are shown as white circles. Fig.13 A map of chromosome 3 to chromosome 6 is shown, with 250k bins.

[0208] In an exemplary embodiment, sequencing data is used to determine the phasing information of polymorphisms known in the starting FFPE sample. For example, sequencing data is used to determine whether certain polymorphisms such as SNPs are present on the same or different DNA molecules. The accuracy of the phasing determined using the method is measured by comparing with the sequence of known sequences such as GIAB samples. For example, in some cases, it was found that between 0-10,000, 132,796 SNPs were found, and 99.059% were in the correct phase. High consistency (>95%) was seen until about 1.5MB (except for 70-80kb bins (which missed 1 of 13) and 1.1-1.3MB bins (which missed 2 of 15)). Within the range of 1.7-1.9MB, 7 of 7 SNP pairs were correctly determined. From these data, the following conclusions can be drawn: despite the low level of pseudo-connection, appropriate long-range information was determined using the FFPE-Chicago method, even reaching the megabase range. Importantly, these “consistency” prediction rates were 95% or higher in many cases, significantly higher than the 50% success rate expected based on random chance.

[0209] Structural phasing information

[0210] At present, structural and phasing analysis (for example, for medical purposes) remains challenging. For example, there is surprising heterogeneity between cancer, individuals with the same type of cancer, or even within the same tumor. Sorting out the cause effect from the result effect may require very high precision and throughput at a low cost for each sample. In the field of personalized medicine, one of the gold standards for genomic therapy is a sequenced genome with all variants fully characterized and phased, including large and small structural rearrangements and novel mutations. In order to achieve this standard, previous technologies require efforts similar to those required for de novo assembly, which is currently too expensive and laborious to become a routine medical procedure.

[0211] The phasing information includes maternal / paternal phasing and tumor / non-tumor phasing information. Tumor / non-tumor phasing can be used to distinguish cancer genome information from somatic cell genome information.

[0212] In some embodiments of the present disclosure, a preserved tissue (e.g., FFPE tissue) from a subject can be provided, and the method can return an assembled genome, an alignment with determined variants (including large structural variants and copy number variants), a phased variant call, or any other analysis. In other embodiments, the methods disclosed herein can directly provide a long-distance read pair library for an individual.

[0213] In various embodiments of the present disclosure, method disclosed herein can generate long ranges separated by large distances and read pairs. The upper limit of this distance can be improved by the ability to collect large-size DNA samples. In some cases, reading can span up to 50,60,70,80,90,100,125,150,175,200,225,250,300,400,500,600,700,800,900,1000,1500,2000,2500,3000,4000,5000kbp or more genome distance. In some instances, reading can span up to 500kbp genome distance. In other instances, reading can span up to 2000kbp genome distance. Method disclosed herein can be integrated and constructed based on standard techniques in molecular biology, and further is very suitable for improving efficiency, specificity and genome coverage.

[0214] In other embodiments, the methods disclosed herein can be used with currently used sequencing technologies. For example, the methods can be used in combination with well-tested and / or widely deployed sequencing instruments. In further embodiments, the methods disclosed herein can be used with techniques and methods derived from currently used sequencing technologies.

[0215] In various embodiments, the present disclosure provides one or more methods disclosed herein, which include the step of detecting the physical layout of chromosomes in a preserved (e.g., FFPE) sample or cell. Examples of techniques for detecting the physical layout of chromosomes by sequencing include "C" class technologies, such as chromosome conformation capture ("3C"), circularized chromosome conformation capture ("4C"), carbon copy chromosome capture ("5C"), and Hi-C-based methods; and ChIP-based methods, such as ChIP-loop, ChIP-PET. These technologies utilize the fixation of chromatin in living cells to consolidate spatial relationships in the nucleus. Subsequent processing and sequencing of the product allow researchers to restore a matrix of neighboring associations between genomic regions. Through further analysis, these associations can be used to generate three-dimensional geometric maps of chromosomes because they are physically arranged in preserved (e.g., FFPE) samples. Such technology describes the discrete spatial organization of chromosomes and provides an accurate view of functional interactions between chromosome sites.

[0216] In some embodiments, intrachromosomal interactions are related to chromosomal connectivity. In some cases, intrachromosomal data can help genome assembly. In some cases, chromatin is reconstructed in vitro. This may be advantageous because chromatin, particularly histones (the main protein component of chromatin), is important for fixation under the most common "C" class technology 3C, 4C, 5C and Hi-C for detecting chromatin conformation and structure by sequencing. Chromatin is highly non-specific with respect to sequence, and will usually be assembled uniformly across the genome. In some cases, the genome of a species that does not use chromatin can be assembled on reconstructed chromatin, thereby expanding the scope of the present disclosure to all life domains.

[0217] Reading pair data can be obtained from chromatin conformation capture technology. In some instances, connection or other markings are completed to mark the genomic region physically close to each other. The crosslinking of the complex can be completed according to the appropriate method further described in detail elsewhere herein or known in the art, so that proteins (such as histones) and DNA molecules in chromatin, such as genomic DNA, are stably combined in the complex. In some cases, by extracting DNA-protein complexes under conditions that make such complexes not degraded, for example, by excluding proteinase K processing, the crosslinking caused by sample preservation (such as by fixing) is utilized. For example, when being a part of a structure such as chromatin, the nucleotide segments that are not close to each other along the genome sequence can be physically close to each other. Such nucleotide segments can be linked together, then analyzed according to the method of the present disclosure. For example, the nucleotide segments connected can be sequenced, and the distance (insertion distance) between the sequencing ends of the segments of two connections can be analyzed. Fig.14A Shown is a graph of the probability of inserts within a particular range versus the insertion distance in base pairs (bp) for preserved samples (eg, FFPE samples) analyzed by the techniques of the present disclosure. Fig. 14B Similar graphs are shown for samples analyzed using the Chicago method. In both graphs, the x-axis shows the insertion distance (bp), from 0 to 300,000, and the y-axis shows the probability of an insertion fragment at that distance, from 10 at the top of the axis. 0 To the bottom of the axis 10 -8 (Logarithm).

[0218] In some cases, two or more nucleotide sequences can be cross-linked via proteins bound to one or more nucleotide sequences. One method is to expose chromatin to ultraviolet irradiation (Gilmour et al., Proc. Nat'l. Acad. Sci. USA 81: 4275-4279, 1984). Other methods, such as chemical or physical (e.g., light) cross-linking, can also be used to cross-link polynucleotide segments. Suitable chemical cross-linking agents include, but are not limited to, formaldehyde and psoralen (Solomon et al., Proc. Natl. Acad. Sci. USA 82: 6470-6474, 1985; Solomon et al., Cell 53: 937-947, 1988). For example, cross-linking can be performed by adding 2% formaldehyde to a mixture containing DNA molecules and chromatin proteins. Other examples of reagents that can be used to cross-link DNA include, but are not limited to, ultraviolet rays, mitomycin C, nitrogen mustard, melphalan, 1,3-butadiene diepoxide, cis-diamine dichloroplatinum (II) and cyclophosphamide. Suitably, the cross-linking agent will form bridges of relatively short distances such as about of cross-links, thereby selecting for close interactions that can be reversed.

[0219] Typically, procedures for probing the physical layout of chromosomes, such as Hi-C-based techniques, utilize chromatin formed in cells / organisms, such as chromatin isolated from cultured cells or primary tissues. The Chicago-based method not only provides the use of such techniques using chromatin isolated from cells / organisms, but also provides the use of such techniques using reconstructed chromatin. Reconstructed chromatin is different from chromatin formed in cells / organisms in a variety of features. First, for many samples, the collection of naked DNA samples can be achieved by using a variety of non-invasive to invasive methods, such as by collecting body fluids, swabbing the oral or rectal area, collecting epithelial samples, etc. Second, reconstructing chromatin substantially prevents the formation of interchromosomal and other long-range interactions, which generate artifacts for genome assembly and unit type phasing. In some cases, according to the methods and compositions of the present disclosure, the sample may have less than about 20%, 15%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.4%, 0.3%, 0.2%, 0.1% or less of interchromosomal or intermolecular crosslinking. In some instances, the sample may have less than about 5% of interchromosomal or intermolecular crosslinking. In some instances, the sample may have less than about 3% of interchromosomal or intermolecular crosslinking. In other instances, there may be less than about 1% of interchromosomal or intermolecular crosslinking. Third, the frequency of sites capable of crosslinking can be adjusted, and therefore the frequency of intramolecular crosslinking within polynucleotides can be adjusted. For example, the ratio of DNA to histones can be varied so that the nucleosome density can be adjusted to a desired value. In some cases, the nucleosome density is reduced to below physiological levels. Therefore, the distribution of crosslinking can be changed to facilitate longer range interactions. In some embodiments, subsamples with different crosslink densities can be prepared to cover short-range and long-range associations. For example, the crosslinking conditions can be adjusted so that at least about 1%, about 2%, about 3%, about 4%, about 5%, about 6%, about 7%, about 8%, about 9%, about 10%, about 11%, about 12%, about 13%, about 14%, about 15%, about 16%, about 17%, about 18%, about 19%, about 20%, about 25%, about 30%, about 40%, about 45%, about 50%, about 60%, about 70%, about 80%, about 90%, about 95%, or about 100% of the crosslink density is at least about 1%. % of cross-links occur between DNA segments that are at least about 50 kb, about 60 kb, about 70 kb, about 80 kb, about 90 kb, about 100 kb, about 110 kb, about 120 kb, about 130 kb, about 140 kb, about 150 kb, about 160 kb, about 180 kb, about 200 kb, about 250 kb, about 300 kb, about 350 kb, about 400 kb, about 450 kb, or about 500 kb apart on the sample DNA molecules.

[0220] The high accuracy required for cancer genome sequencing can be achieved using the methods and systems described herein. When sequencing cancer genomes, inaccurate reference genomes may bring challenges to base determination. Heterogeneous samples and small starting materials, such as samples obtained by biopsy, introduce additional challenges. In addition, the detection of large-scale structural variants and / or loss of heterozygosity is often crucial for cancer genome sequencing, as well as the ability to distinguish somatic variants from base determination errors.

[0221] The systems and methods described herein can generate accurate long sequences from complex samples containing 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, 15, 20 or more different genomes. Mixed samples of normal, benign and / or tumor origin can be analyzed (optionally without normal controls). In some embodiments, accurate long sequences are generated using only 100ng or even only hundreds of genome equivalent starting samples. The systems and methods described herein can allow detection of copy number variants, large-scale structural variants and rearrangements, and phased variant determinations can be obtained on long sequences of nucleotides spanning about 1kbp, about 2kbp, about 5kbp, about 10kbp, 20kbp, about 50kbp, about 100kbp, about 200kbp, about 500kbp, about 1Mbp, about 2Mbp, about 5Mbp, about 10Mbp, about 20Mbp, about 50Mbp or about 100Mbp or more. For example, phase variant determinations can be obtained on long sequences spanning about 1Mbp or about 2Mbp.

[0222] The unit type determined using the methods and systems described herein can be assigned to computing resources, such as computing resources via a network, such as a cloud system. If necessary, the relevant information stored in the computing resources can be used to correct the short variant determination. Structural variants can be detected based on the composite information determined from the short variant and the information stored in the computing resources. The problematic parts of the genome, such as segmental duplications, regions with structural variation tendencies, highly variable medically relevant MHC regions, centromere and telomeric regions, and other heterochromatic regions (including but not limited to those heterochromatic regions with repeated regions, low sequence accuracy, high mutation rates, ALU repeats, segmental duplications, or any other relevant problematic parts known in the art), can be reassembled to improve accuracy.

[0223] Sample types can be assigned to sequence information in local or networked computing resources (such as clouds). Where the source of the information is known, for example, when the source of the information is from cancer or normal tissue, the source can be assigned to the sample as part of the sample type. Other sample type examples typically include, but are not limited to, tissue type, sample collection method, presence of infection, type of infection, treatment method, size of the sample, etc. Where a complete or partial comparative genome sequence is available, such as a comparison of a normal genome with a cancer genome, the difference between the sample data and the comparative genome sequence can be determined and optionally output.

[0224] Haplotype phasing methods

[0225] Since the read pairs generated by the method disclosed herein are generally derived from intrachromosomal contacts, any read pairs containing heterozygous sites will also carry information about their phasing. Using this information, reliable phasing of short distances, medium distances, and even long distances (megabases) can be performed quickly and accurately. Experiments designed to phase the data from one of 1000 genome triplets (a collection of mother / father / offspring genomes) have reliably inferred phasing. In addition, the cell type reconstruction using adjacent connections similar to Selvaraj et al. (Nature Biotechnology 31: 1111-1118 (2013)) can also be used together with the cell type phasing method disclosed herein.

[0226] For example, the cell type reconstruction using the method based on the neighbor connection can also be used for genome phasing in the method disclosed herein. The cell type reconstruction using the method based on the neighbor connection combines the neighbor connection and DNA sequencing with the probability algorithm of the cell type assembly. First, the chromosome capture scheme such as the Hi-C scheme is used for the neighbor connection sequencing. These methods can capture DNA fragments from two distant genomic loci that are looped in three-dimensional space. After the obtained DNA library is subjected to shotgun DNA sequencing, the paired end sequencing reads have an "insert size" ranging from hundreds of base pairs to tens of millions of base pairs. Therefore, the short DNA fragments generated in the Hi-C experiment can produce small cell type blocks, and the long fragments can eventually connect these small blocks together. In the case of sufficient sequencing coverage, this method can connect variants in discontinuous blocks, and assemble each such block into a single cell type. Then the data is combined with the probability algorithm for cell type assembly. The probability algorithm uses a graph in which nodes correspond to heterozygous variants and edges correspond to overlapping sequence fragments of connectable variants. The graph may contain false edges caused by sequencing errors or trans interactions. The maximum cut algorithm is then used to predict the parsimonious solution with the greatest consistency with the haplotype information provided by the input sequencing read set. Since neighbor joining generates a larger graph than conventional genome sequencing or paired sequencing, the computation time and number of iterations are modified so that haplotypes can be predicted at reasonable speed and with high accuracy. The resulting data can then be used to guide local phasing using Beagle software and sequencing data from the genome project to generate haplotypes across chromosomes with high resolution and accuracy.

[0227] Determining phase information using paired-end

[0228] Further provided herein are methods and compositions for determining phase information from paired ends derived from FFPE samples. Paired ends can be generated by any disclosed method or by the method further described in the provided embodiments. For example, in the case where a DNA molecule is bound to a solid surface and subsequently cut, after reconnecting the free ends, the reconnected DNA segments are released from the solid-phase attached DNA molecules, such as by restriction digestion. The release produces multiple paired end fragments. In some cases, paired ends are connected to amplification adapters, amplified, and sequenced with short read technology. In these cases, paired ends from multiple different solid-phase bound DNA molecules are within the sample sequenced. However, it is confidently concluded that for either side of the paired end joint, the joint adjacent sequence is derived from the common phase of the common molecule. In the case where the paired end is connected to a point-break oligonucleotide, the paired end joint in the sequencing read is identified by the point-break oligonucleotide sequence. In other cases, the paired ends are connected by modified nucleotides, and the ends can be identified based on the sequence of the modified nucleotides used.

[0229] Alternatively, after releasing the paired ends, the free paired ends are connected to the amplification adapter and amplified. In these cases, multiple paired ends are then connected in batches to generate long molecules read using long read sequencing technology. In other examples, the released paired ends are connected in batches to each other without an intermediate amplification step. In either case, the embedded read pair can be identified via a natural DNA sequence adjacent to a connection sequence (such as a point-break sequence or a modified nucleotide). The multiplexed paired ends are read on a long sequence device, and the sequence information of multiple joints is obtained. Since the paired ends are derived from a variety of different solid-phase-bound DNA molecules, it is found that the sequences spanning two separate paired ends, such as those flanking the amplification adapter sequence, are mapped to a variety of different DNA molecules. However, it is confidently concluded that for either side of the paired end joint, the joint adjacent sequence is derived from the common phase of the common molecule. For example, in the case of paired ends from a point-break molecule, the sequence flanking the point-break sequence is confidently assigned to a common DNA molecule. In a preferred case, because each paired end is multiplexed using the methods and compositions disclosed herein, multiple paired ends can be sequenced in a single read.

[0230] In preferred embodiments, sequencing data generated using the methods and compositions described herein are used to generate phased de novo sequence assemblies, determine phase information, and / or identify structural variations.

[0231] Identify structural variants and other genetic features

[0232] See also Fig.15A and Fig. 15B , provides examples of mapping positions of read pairs on a reference sequence such as GRCh38, which are generated by proximity ligation of DNA from reassembled chromatin and plotted near structural differences between GM12878 and the reference. Each read pair generated is represented above and below the diagonal line. Above the diagonal line, the shading represents the map quality score on the scale shown; below the diagonal line, the shading represents the inferred unit type phase of the generated read pair based on the overlap with the phased SNP. In some embodiments, the generated map depicts an inversion with a flanking repeat region, such as Fig. 15B In some embodiments, the generated graph depicts the data of the phased heterozygous deletion, such as Fig. 15B as shown in .

[0233] Mapping paired sequence reads from one individual to a reference is the most commonly used sequence-based method for identifying differences in contiguous nucleic acid or genomic structure such as inversions, deletions, and duplications (Tuzun et al., 2005). Fig.15A and Fig. 15BShown how to read the DNA of the reassembled chromatin from GM12878 by adjacent connection and generate and map to the human reference genome GRCh38, to reveal two such structural differences.In order to estimate the sensitivity and specificity of the read pair data for identifying structural differences, the maximum likelihood discriminator of the simulated data set constructed for simulating the influence of heterozygous inversion is tested.By reading and mapping to the GRCh38 reference sequence from the generated NA12878, the interval of bounding length L is randomly selected, and each generated read pair is randomly and independently assigned to inversion or reference unit type, and the coordinates of mapping are edited accordingly to construct test data.Non-allelic homologous recombination is the cause of many structural variations observed in the human genome, resulting in many variation breakpoints (Kidd et al., 2008) occurring in long blocks of repetitive sequences.The influence of repetitive sequences of different lengths around inversion breakpoints is simulated by removing all reads mapped within distance W. In the absence of repetitive sequences at the inversion breakpoints, the sensitivity (specificity) was 0.76 (0.88), 0.89 (0.89), and 0.97 (0.94) for 1 Kbp, 2 Kbp, and 5 Kbp inversions, respectively. When repetitive (unmappable) sequences in the 1 Kbp region at the inversion breakpoint were used in the simulations, the sensitivity (specificity) for the 5 Kbp inversion was 0.81 (0.76).

[0234] performance

[0235] The analysis performed using the techniques disclosed herein can be performed with high accuracy. The analysis can be performed with an accuracy of at least about 50%, 60%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.9%, 99.99%, 99.999% or more. The analysis can be performed with an accuracy of at least 70%. The analysis can be performed with an accuracy of at least 80%. The analysis can be performed with an accuracy of at least 90%.

[0236] The analysis performed using the techniques disclosed herein can be performed with high specificity. The analysis can be performed with a specificity of at least about 50%, 60%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.9%, 99.99%, 99.999% or more. The analysis can be performed with a specificity of at least 70%. The analysis can be performed with a specificity of at least 80%. The analysis can be performed with a specificity of at least 90%.

[0237] The analysis performed using the techniques disclosed herein can be performed with high sensitivity. The analysis can be performed with a sensitivity of at least about 50%, 60%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, 99.9%, 99.99%, 99.999% or more. The analysis can be performed with a sensitivity of at least 70%. The analysis can be performed with a sensitivity of at least 80%. The analysis can be performed with a sensitivity of at least 90%.

[0238] The use of the techniques of the present disclosure can improve the functionality of a computer system used to implement the techniques. For example, the techniques can reduce the processing time of a given analysis by at least about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or more. The techniques can reduce the memory requirements of a given analysis by at least about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or more.

[0239] The use of the techniques of the present disclosure may enable analyses that were previously not possible. For example, certain genetic features may be detected from sequence information that would not have been detected from such information if the methods of the present disclosure were not used.

[0240] Machine Learning

[0241] A variety of techniques can be used to analyze to identify features such as contacts and rearrangements (including but not limited to deletions, duplications, insertions, inversions or reversals, translocations, joining, fusions and splits) and other interactions. Analysis techniques can include statistical and probabilistic analysis, signal processing including Fourier analysis, computer vision and other image processing, language processing (e.g., natural language processing) and machine learning. For example, data configurations indicating features such as the above features in interaction graphs such as contact matrices can be analyzed. In some cases, filters can be applied to graphs or other data. The filter can be a convolution filter, including but not limited to a smoothing filter (e.g., kernel smoothing or Savitzky-Golay filter, Gaussian blur, etc.).

[0242] Some embodiments involve machine learning as a component of genome structure determination, so some computer systems are configured to include a module with machine learning capabilities. The machine learning module includes at least one of the modalities listed below to constitute a machine learning function.

[0243] The modalities that make up machine learning exhibit data filtering capabilities in various ways so that automated mass spectrometry data point detection and interpretation can be performed. In some cases, this modality is facilitated by the presence of predictive patterns that indicate various genomic structural changes, such as inversions, insertions, deletions, or translocations.

[0244] The modalities that make up machine learning exhibit data processing or data manipulation capabilities in various ways to put read pair frequencies in a form that is conducive to downstream analysis. Examples of data manipulation include, but are not necessarily limited to, logarithmic transformation, assigning scaling ratios, or mapping the data to carefully designed features to put the data in a form that is conducive to downstream analysis.

[0245] A machine learning data analysis component as disclosed herein periodically processes a read pair data set of numerous features, such as 1 to 10,000 features or 2 to 300,000 features, or a number within or above any of these ranges. In some cases, the data analysis involves at least 1k, 2k, 3k, 4k, 5k, 6k, 7k, 8k, 9k, 10k, 20k, 30k, 40k, 50k, 60k, 70k, 80k, 90k, 100k, 120k, 140k, 160k, 180k, 200k, 220k, 2240k, 260k, 280k, 300k or more than 300k features.

[0246] The read pair distribution pattern is determined using any number of methods consistent with the disclosure herein. In some cases, the read pair distribution pattern selection includes elastic net, information gain, random forest input, or other feature selection methods consistent with the disclosure herein and familiar to those skilled in the art.

[0247] Again using any number of methods consistent with the disclosure herein, the selected read pair distribution pattern is matched with a predicted pattern indicative of a genomic structural change. In some cases, read pair pattern detection comprises logistic regression, SVM, random forest, KNN, or other classifier methods consistent with the disclosure herein and familiar to those skilled in the art.

[0248] Applying machine learning or providing a machine learning module on a computer configured for the analyses disclosed herein allows for the detection of genomic structural changes that are meaningful for asymptomatic disease detection or early detection as part of an ongoing monitoring program to identify a disease or condition before symptoms appear or when intervention is more achievable or more likely to result in a successful outcome.

[0249] Applying machine learning or providing a machine learning module on a computer configured for the analyses disclosed herein also allows for the identification of structural rearrangements in individuals undergoing drug treatment, for example as part of a drug trial, so that trial results for individuals or populations can be correlated concurrently or retrospectively to identify specific genomic structural events that are positively or negatively correlated with drug efficacy.

[0250] Applying machine learning or providing a machine learning module on a computer configured for the analysis disclosed herein also allows identification of structural rearrangements corresponding to specific regions of a genetically heterogeneous sample (e.g., a tumor tissue sample collected without homogenization) so as to preserve positional information in the sample. Since some tumor regions are known to correspond to cell populations that are particularly susceptible to metastasis or tumor spread, identification of genomic rearrangements or other phase information associated with such cell populations can help select treatment options for these particularly dangerous cell populations.

[0251] Monitoring is often, but not necessarily, conducted in conjunction with or in support of a genetic assessment that indicates a genetic susceptibility to a condition whose onset or progression characteristics are monitored. Similarly, in some cases, machine learning is used to facilitate monitoring or assessment of the therapeutic efficacy of a treatment regimen so that the treatment regimen can be modified, continued, or addressed over time, as indicated by ongoing proteomics-mediated monitoring.

[0252] Machine learning methods and computer systems having modules configured to execute machine learning algorithms facilitate identification of phase information or genomic rearrangements in data sets of varying complexity. In some cases, phase information or genomic rearrangements are identified from an untargeted database comprising a large amount of mass spectrometry data, such as data obtained from a single individual at multiple time points, samples collected from multiple individuals (such as multiple individuals with known status or known final treatment outcomes or responses for a condition of interest), or data from multiple time points and multiple individuals.

[0253] Alternatively, in some cases, machine learning facilitates refinement of genomic rearrangement or phase information by analyzing a database against which the genomic rearrangement or phase information is directed, for example, by collecting genomic rearrangement or phase information from a single individual at multiple time points (when the individual's health status is known for that time point), or by collecting sequence information from multiple individuals with known status for a condition of interest, or by collecting sequence information from multiple individuals at multiple time points. Obviously, in some cases, the collection of sequence information is facilitated by using preserved samples, such as cross-linked samples collected from surgery or FFPE samples collected from drug trials.

[0254] Thus, sequence information is collected alone or in combination with information on drug trial results or surgical intervention results. The sequence data is subjected to machine learning, for example on a computer system configured as disclosed herein, to identify a subset of read pairs that indicate a pattern corresponding to a genomic rearrangement, which alone or in combination with one or more additional markers illustrates a health status signal. Thus, in some cases, machine learning helps identify sequences - DNA or RNA sequences, or genomic rearrangements that alone provide information about an individual's health status.

[0255] An exemplary machine learning method consistent with the above disclosure is a convolutional neural network (CNN). CNN can be used, for example, to classify positive samples and negative samples. An exemplary CNN architecture comprises 2 fully connected convolutional hidden layers, each of which is followed by a maximum pooling layer of a certain number of neurons and a final output layer, such as a number of neurons that can only be divided by 2 or a factor of 2, such as 128, 256, 512, 1024 other numbers or neurons with logit activation functions. In alternative embodiments, a wide range of neuron numbers are compatible with the disclosure herein, such a number is less than 50, 50, 60, 64, 70, 80, 90, 100, 120, 140, 160, 180, 200, 250, 300, 350, 400, 450, 500, 550, 600, 650, 700, 750, 800, 850 , 900, 950, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2048, 2100, 2200, 2300, 2400, 2500, 2600, 2700, 2800, 2900, 3000 or more within the range defined by the endpoints of 3000 or more.

[0256] In some implementations of machine learning such as CNN, the training data uses read pair count information, and the intra-chromosomal matrix is ​​normalized using, for example, the inverse of the distance from the diagonal to the read pair mapping point. Alternatively or in combination, other parameters (such as reference mappability, restriction site distribution or other parameters) are used as additional channels to create a multi-channel neural network (such as a CNN network).

[0257] Image classification is achieved using feature localization through many state-of-the-art networks such as YOLO, Mask R-CNN, Fast R-CNN, and others. Alternatively, a domain-specific architecture is designed for a specific application.

[0258] Computer Systems

[0259] Fig.18AA computer system 401 is shown that is programmed or otherwise configured to implement the methods provided herein. The computer system 401 may be a user's electronic device or a computer system remotely located relative to the electronic device. The electronic device may be a mobile electronic device.

[0260] Computer system 401 includes a central processing unit (CPU, also referred to herein as a "processor" and "computer processor") 405, which may be a single-core or multi-core processor, or multiple processors for parallel processing. Computer system 401 also includes memory or storage location 410 (e.g., random access memory, read-only memory, flash memory), an electronic storage unit 415 (e.g., a hard disk), a communication interface 420 (e.g., a network adapter) for communicating with one or more other systems, and peripherals 425, such as cache memory, other memory, data storage, and / or an electronic display adapter. Memory 410, storage unit 415, interface 420, and peripherals 425 communicate with CPU 405 via a communication bus (solid lines) such as a motherboard. Storage unit 415 may be a data storage unit (or data repository) for storing data. Computer system 401 may be operably coupled to a computer network ("network") 430 by means of communication interface 420. Network 430 can be the Internet, the Internet and / or an extranet, and / or an intranet and / or an extranet in communication with the Internet. In some cases, network 430 is a telecommunication and / or data network. Network 430 can include one or more computer servers that can implement distributed computing, such as cloud computing. In some cases, with the aid of computer system 401, network 430 can implement a peer-to-peer network, which can enable devices coupled to computer system 401 to act as clients or servers.

[0261] The CPU 405 may execute a series of machine-readable instructions, which may be embodied in a program or software. The instructions may be stored in a storage location, such as a memory 410. The instructions may be directed to the CPU 405, which may then program or otherwise configure the CPU 405 to implement the methods of the present disclosure. Examples of operations performed by the CPU 405 may include fetching, decoding, executing, and writing back.

[0262] CPU 405 may be part of a circuit such as an integrated circuit. One or more other components of system 401 may be included in the circuit. In some cases, the circuit is an application specific integrated circuit (ASIC).

[0263] Storage unit 415 can store files, such as drivers, libraries, and saved programs. Storage unit 415 can store user data, such as user preferences and user programs. In some cases, computer system 401 can include one or more additional data storage units that are external to computer system 401, such as located on a remote server that communicates with computer system 401 via an intranet or the Internet.

[0264] Computer system 401 may communicate with one or more remote computer systems via network 430. For example, computer system 401 may communicate with a remote computer system of a user (e.g., a service provider). Examples of remote computer systems include personal computers (e.g., portable PCs), tablet or tablet-type PCs (e.g., iPad, GalaxyTab), phones, smartphones (e.g. iPhone, Android-enabled devices, ) or a personal digital assistant. A user can access the computer system 401 via the network 430.

[0265] Methods as described herein may be implemented by machine (e.g., computer processor) executable code stored on an electronic storage location of computer system 401, such as on memory 410 or electronic storage unit 415. The machine executable code or machine readable code may be provided in the form of software.

[0266] During use, the code may be executed by processor 405. In some cases, the code may be retrieved from storage unit 415 and stored on memory 410 for access by processor 1005. In some cases, electronic storage unit 415 may be excluded and machine-executable instructions may be stored on memory 410.

[0267] The code may be precompiled and configured for use with a machine having a processor suitable for executing the code, or may be compiled during runtime. The code may be provided in a programming language, which may be selected to enable the code to be executed in a precompiled or as-compiled manner.

[0268] Aspects of the systems and methods provided herein, such as computer system 1001, may be embodied in programming. Aspects of the present technology may be considered as "products" or "articles of manufacture", which are generally in the form of machine (or processor) executable code and / or related data carried or embodied on a certain type of machine-readable medium. The machine executable code may be stored in an electronic storage unit such as a memory (e.g., read-only memory, random access memory, flash memory) or a hard disk. "Storage" type media may include any or all tangible memories, processors, etc. of a computer, or related modules thereof, such as various semiconductor memories, tape drives, disk drives, etc., which may provide non-temporary storage for software programming at any time. All or part of the software may sometimes communicate via the Internet or various other telecommunications networks. Such communication, for example, may enable software to be loaded from one computer or processor to another, for example, from a management server or host to a computer platform of an application server. Therefore, another type of medium that can carry software elements includes optical waves, radio waves, and electromagnetic waves, such as across physical interfaces between local devices, through wired and optical landline networks, and through various air links. Physical elements that carry such waves, such as wired or wireless links, optical links, etc., can also be considered as media that carry the software. As used herein, unless restricted to non-transitory tangible "storage" media, terms such as computer or machine "readable media" refer to any medium that participates in providing instructions to a processor for execution.

[0269] Thus, machine-readable media such as computer executable code may take many forms, including but not limited to tangible storage media, carrier media, or physical transmission media. Non-volatile storage media include, for example, optical or magnetic disks, such as any storage device in any computer, and the like, such as may be used to implement the databases shown in the accompanying drawings, and the like. Volatile storage media include dynamic memory, such as the main memory of such a computer platform. Tangible transmission media include: coaxial cables, copper wire, and optical fibers, including wires that comprise a bus within a computer system. Carrier transmission media may take the form of electrical or electromagnetic signals, or acoustic or light waves, such as those generated during radio frequency (RF) and infrared (IR) data communications. Thus, common forms of computer readable media include, for example, a floppy disk, a flexible disk, a hard disk, a magnetic tape, any other magnetic medium, a CD-ROM, a DVD or DVD-ROM, any other optical medium, a punched card stock tape, any other physical storage medium having a pattern of holes, a RAM, a ROM, a PROM and an EPROM, a FLASH-EPROM, any other memory chip or cartridge, a carrier wave that transmits data or instructions, a cable or link that transmits such a carrier wave, or any other medium from which a computer can read programming code and / or data. Many of these forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to a processor for execution.

[0270] Computer system 401 may include or be in communication with electronic display 435 including user interface (UI) 440 for providing output and readouts of, for example, trained algorithms. Examples of UIs include, but are not limited to, graphical user interfaces (GUIs) and web-based user interfaces.

[0271] The methods and systems of the present disclosure may be implemented by one or more algorithms. The algorithms may be implemented by software when executed by the central processing unit 405 .

[0272] In some cases, the computer systems herein are configured to perform machine learning operations, such as those disclosed in the specification herein or known to those skilled in the art.

[0273] Fig.18B The computer system 600 shown in the figure can be understood as a logical device that can read instructions from a medium 611 and / or a network port 605, which can be optionally connected to a server 609 having a fixed medium 612. The system (such as Fig.18B601, a disk drive 603, an optional input device such as a keyboard 615 and / or a mouse 616, and an optional monitor 607. Data communication with a server at a local or remote location can be achieved through the indicated communication medium. The communication medium may include any means of transmitting and / or receiving data. For example, the communication medium may be a network connection, a wireless connection, or an Internet connection. Such a connection may provide communication through the World Wide Web. It is contemplated that data related to the present disclosure may be transmitted through such a network or such as Fig.18B The connection shown is received and / or inspected by party 622 for transmission.

[0274] Fig.18C 7 is a block diagram illustrating a first exemplary architecture of a computer system 700 that may be used in conjunction with the exemplary embodiments described herein. Fig.18C , an exemplary computer system includes a processor 702 for processing instructions. Non-limiting examples of processors include: Intel Xeon™ processor, AMD Opteron™ processor, Samsung 32-bitRISC ARM1176JZ (F) -S v1.0™ processor, ARM Cortex-A8 Samsung S5PC100™ processor, ARMCortex-A8 Apple A4™ processor, Marvell PXA 930™ processor or a functionally equivalent processor. Multiple execution threads can be used for parallel processing. In some embodiments, multiple processors or processors with multiple cores are used, whether in a single computer system, in a cluster, or across a network including multiple computers, cellular phones and / or personal data assistant devices.

[0275] like Fig.18C As shown in , cache memory 704 may be connected or incorporated into processor 702 to provide high-speed storage for instructions or data that are recently or frequently used by processor 702. Processor 702 is connected to north bridge 706 via processor bus 708. North bridge 706 is connected to random access memory (RAM) 710 via memory bus 712, and manages access to RAM 710 by processor 702. North bridge 706 is also connected to south bridge 714 via chipset bus 716. South bridge 714 is in turn connected to peripheral bus 718. The peripheral bus may be, for example, PCI, PCI-X, PCI Express or other peripheral bus. North bridge and south bridge are generally referred to as processor chipsets, and manage data transfer between processor, RAM and peripheral components on peripheral bus 718. In some alternative architectures, the function of north bridge may be incorporated into the processor without using a separate north bridge chip.

[0276] In some embodiments, the system 700 includes an accelerator card 722 attached to the peripheral bus 718. The accelerator may include a field programmable gate array (FPGA) or other hardware for accelerating certain processes. For example, the accelerator may be used for adaptive data reconstruction or to evaluate algebraic expressions used in extended set processing.

[0277] Software and data are stored in external storage 724 and may be loaded into RAM 710 and / or cache 704 for use by the processor. System 2000 includes an operating system for managing system resources; non-limiting examples of operating systems include: Linux, Windows TM 、MACOS TM 、BlackBerry OS TM 、iOS TM An operating system and other functionally equivalent operating systems, and application software running on the operating system for managing data storage and optimization according to exemplary embodiments of the present invention.

[0278] In this example, system 700 also includes network interface cards (NICs) 720 and 721 connected to the peripheral bus for providing a network interface to external storage such as network attached storage (NAS) and other computer systems that can be used for distributed parallel processing.

[0279] Fig.18D A schematic diagram of a network 2100 is shown with multiple computer systems 2102a and 2102b, multiple cell phones and personal data assistants 2102c, and network attached storage (NAS) 2104a and 2104b. In an exemplary embodiment, systems 2102a, 2102b, and 2102c can manage data storage and optimize data access to data stored in network attached storage (NAS) 2104a and 2104b. Mathematical models can be used for data and evaluated using distributed parallel processing in computer systems 2102a and 2102b and cell phones and personal data assistant systems 2102c. Computer systems 2102a and 2102b and cell phones and personal data assistant systems 2102c can also provide parallel processing for adaptive data reconstruction of data stored in network attached storage (NAS) 2104a and 2104a. Fig.18D Only examples are shown, and a variety of other computer architectures and systems may be used in conjunction with various embodiments of the present invention. For example, blade servers may be used to provide parallel processing. Processor blades may be connected via a backplane to provide parallel processing. Storage may also be connected to the backplane via a separate network interface or as a network attached storage (NAS).

[0280] In some exemplary embodiments, a processor may maintain a separate memory space and transmit data through a network interface, backplane or other connector for parallel processing by other processors. In other embodiments, some or all processors may use a shared virtual address memory space.

[0281] Fig.18E 900 is a block diagram of a multiprocessor computer system 900 using a shared virtual address memory space according to an exemplary embodiment. The system includes multiple processors 902a-f that can access a shared memory subsystem 904. The system includes multiple programmable hardware memory algorithm processors (MAPs) 906a-f in the memory subsystem 904. Each MAP 906a-f can include memory 908a-f and one or more field programmable gate arrays (FPGAs) 910a-f. The MAP provides a configurable functional unit, and a specific algorithm or algorithm portion can be provided to the FPGA 910a-f for processing in close coordination with the corresponding processor. For example, in an exemplary embodiment, the MAP can be used to evaluate algebraic expressions about data models and perform adaptive data reorganization. In this example, to achieve these purposes, each MAP can be globally accessible to all processors. In one configuration, each MAP can use direct memory access (DMA) to access the associated memory 908a-f, so that it can perform tasks independently of the corresponding microprocessor 902a-f and asynchronously with the microprocessor. In this configuration, a MAP can provide results directly to another MAP for pipelining and parallel execution of algorithms.

[0282] The computer architectures and systems described above are exemplary only, and a variety of other computer, cell phone, and personal data assistant architectures and systems may be used in conjunction with the exemplary embodiments, including systems using any combination of general purpose processors, coprocessors, FPGAs and other programmable logic devices, systems on a chip (SOCs), application specific integrated circuits (ASICs), and other processing and logic elements. In some embodiments, all or part of the computer system may be implemented in software or hardware. Any type of data storage media may be used in conjunction with the exemplary embodiments, including random access memory, hard drives, flash memory, tape drives, disk arrays, network attached storage (NAS), and other local or distributed data storage devices and systems.

[0283] In an exemplary embodiment, a computer system may be implemented using software modules executed on any of the above or other computer architectures and systems. In other embodiments, the functionality of the system may be implemented partially or completely in firmware, programmable logic devices such as Fig.18EThe invention is implemented in a field programmable gate array (FPGA), a system on a chip (SOC), an application specific integrated circuit (ASIC), or other processing and logic elements as shown.

[0284] The methods and systems disclosed herein provide numerous advantages over methods in use at the time of filing this application.

[0285] Some methods and computing systems disclosed herein cluster contigs in a manner that is independent of the number of chromosomes of the organism. A more conservative threshold for contig-contig connections with respect to single-connection clusters is applied to assemble the resulting smaller contig clusters into scaffolds, and subsequent splicing connections can be performed by a variety of methods disclosed herein.

[0286] In some embodiments, the method disclosed herein does not involve clustering substantially, but directly performs a tree generation step, followed by topological tree pruning. In some embodiments, more than one clustering method may be used, such as the Markov clustering algorithm (MCL algorithm). Without being limited by theory, topological pruning may be performed to prevent misassembly by particularly carefully handling these edges and avoiding assembly misconnections.

[0287] After the contig order in the fixed support, the contig orientation can be optimized by using a dynamic programming algorithm. Such a method of only reading contig pairs that are mapped to adjacent contig pairs in the sorting helps to optimize the score without considering any contigs that are shorter than the maximum interval of the external and unassembled better fragment pairs. In order to improve the orientation step, in addition to the nearest contig score interactions, contigs that are not the nearest contig score interactions can also be considered by using an algorithm that merges data from all pairs of contig pairs mapped to a maximum of w-2 intermediate contigs, for example using two or more contigs in the sorting, such as 2, 3, 4, 5, 6, 7, 8, 9, 10 or more than 10 contig values.

[0288] In some embodiments, the accuracy of the insertion step can be improved. Without being bound by any theory, after creating the backbone, in the assembly of contigs with shorter than the maximum spacing between good read pairs, data from contigs within the neighborhood of the w contigs along the sort are included when excluding contigs from the backbone and reinserting them at locations that maximize the amount of connectivity between adjacent contigs.

[0289] In some other embodiments, the orientation step can be improved by considering the interaction of superimposed group scores exceeding the nearest neighbor. After the superimposed group order in the fixed support, the superimposed group direction is optimized by using a dynamic programming algorithm. Only the read pairs mapped to the adjacent superimposed group pairs in the sorting contribute to the optimization of the score. In some cases, an algorithm that merges all pairs of data from superimposed group pairs mapped to the maximum w-2 intermediate superimposed groups in the sorting can be used to assemble any superimposed group with a maximum interval shorter than the better fragment pair. For example, two or more superimposed groups in the sorting are used, such as 2, 3, 4, 5, 6, 7, 8, 9, 10 or more than 10 values.

[0290] In some embodiments, the sorting and orientation accuracy can be improved by more tightly integrating the sorting and orientation steps. An initial graph can be constructed so that the nodes in the graph are the ends of contigs, and the two end nodes of each contig are connected by edges. The log-likelihood ratio scores of the edges between contigs are calculated under the assumption of a specific short gap size, and then sorted. As the edge scores of the list are operated in descending order, new edges are accepted or rejected depending on whether they will increase or decrease the total score of the assembly. It is worth noting that even edges with positive scores can reduce the sum of the scores of the contigs in the assembly, because accepting edges (meaning inserting one or more contigs into the gap of an existing support) will increase the gap size between the connected contig pairs on either side of the gap, which may give them a lower score.

[0291] In addition, the maximum likelihood gap size can be calculated efficiently. The overall accuracy of the reported assembly can be increased by estimating the length of unknown sequence between consecutive contigs. Assuming a model of the library creation process that includes a model probability density function (PDF) of the spacing d between library read pairs, the spacing d between pairs that span the gap is calculated by i The maximum likelihood gap length can be found by maximizing the combined likelihood of . For differentiable model PDFs, efficient iterative optimization methods (e.g., Newton-Raphson) can be used.

[0292] The elements of the methods and compositions disclosed herein are to assemble the stacks into configurations that are locally optimal for the stack order, orientation, or order and orientation, for example, in a stack window of 2, 3, 4, 5, 6 or more than 6 stacks, while being executable or obtainable within a relatively short amount of time (such as 8, 7, 6, 5, 4, 3, 2 hours or less than 2 hours). Therefore, in some cases, the methods herein allow a high degree of computing power to be brought to computationally intensive problems without using a large amount of computing time and without the need to explore a very large global computing space. On the contrary, in most cases, local sorting achieves the appropriate and accurate sorting of the stacks, and then the computational intensity is expended to optimize the local windows of the stacks, rather than optimizing all stacks at once overall. In some cases, configuration optimization is completed within 8, 7, 6, 5, 4, 3, 2 hours or less than 2 hours by using a window size within the range of 3, 4, 5 or 6. For larger window sizes, configuration optimization is completed within a few days to a week.

[0293] Digital processing equipment

[0294] In some embodiments, the stack assembly method described herein includes a digital processing device or its use. In other embodiments, the digital processing device includes one or more hardware central processing units (CPUs) that perform the functions of the device. In other embodiments, the digital processing device further includes an operating system configured to execute executable instructions. In some embodiments, the digital processing device is optionally connected to a computer network. In other embodiments, the digital processing device is optionally connected to the Internet so that it accesses the World Wide Web. In other embodiments, the digital processing device is optionally connected to a cloud computing infrastructure. In other embodiments, the digital processing device is optionally connected to an intranet. In other embodiments, the digital processing device is optionally connected to a data storage device.

[0295] According to the description herein, for non-limiting examples, suitable digital processing devices include server computers, desktop computers, laptop computers, notebook computers, small notebook computers, netbook computers, netpad computers, set-top computers, streaming media devices, handheld computers, Internet application appliances, mobile smart phones, tablet computers, personal digital assistants, video game consoles and vehicles. Those skilled in the art will recognize that many smart phones are suitable for use in the systems described herein. Those skilled in the art will also recognize that televisions, video players and digital music players with optional computer network connectivity are selected to be suitable for use in the systems described herein. Suitable tablet computers include tablet computers with manual, tablet and convertible configurations known to those skilled in the art.

[0296] In some embodiments, the digital processing device includes an operating system configured to execute executable instructions. An operating system is software, such as programs and data, that manages the hardware of the device and provides services for executing applications. Those skilled in the art will recognize that suitable server operating systems include, by way of non-limiting example, FreeBSD, OpenBSD, Linux, Mac OS X Windows and Those skilled in the art will recognize that suitable personal computer operating systems include, by way of non-limiting example, Mac OS and UNIX-like operating systems such as In some embodiments, the operating system is provided by cloud computing. Those skilled in the art will also recognize that suitable mobile smartphone operating systems include, by way of non-limiting example, OS, Research In BlackBerry Windows OS, Windows OS, and

[0297] In some embodiments, the device includes a storage and / or memory device. The storage and / or memory device is one or more physical devices for temporarily or permanently storing data or programs. In some embodiments, the device is a volatile memory and requires power to maintain the stored information. In some embodiments, the device is a non-volatile memory and retains the stored information when the digital processing device is not powered. In other embodiments, the non-volatile memory includes a flash memory. In some embodiments, the non-volatile memory includes a dynamic random access memory (DRAM). In some embodiments, the non-volatile memory includes a ferroelectric random access memory (FRAM). In some embodiments, the non-volatile memory includes a phase change random access memory (PRAM). Optionally, the device is a storage device, including, by way of non-limiting example, a CD-ROM, a DVD, a flash memory device, a disk drive, a tape drive, an optical drive, and cloud computing-based storage. In other embodiments, the storage and / or memory device is a combination of devices such as those disclosed herein.

[0298] Some digital processing devices include a display that transmits visual information to a user, such as a cathode ray tube (CRT), a liquid crystal display (LCD), a thin film transistor liquid crystal display (TFT-LCD), an organic light emitting diode (OLED) display such as a passive matrix OLED (PMOLED) or an active matrix OLED (AMOLED) display, a plasma display, a video projector, or a combination of devices such as those disclosed herein.

[0299] Typically, a digital processing device includes an input device such as a keyboard, a pointing device, including, by way of non-limiting example, a mouse, a trackball, a trackpad, a joystick, a game controller, or a stylus, that receives information from a user. In some embodiments, the input device is a touch screen or multi-touch screen, a microphone that captures voice or other sound input, or a camera or other sensor that captures motion or visual input. In other embodiments, the input device is a Kinect, Leap Motion, etc. Typically, the input device is a combination of devices such as those disclosed herein.

[0300] Non-transitory computer readable storage medium

[0301] In some embodiments, the stack assembly method disclosed herein involves one or more non-transitory computer-readable storage media, which are encoded with a program containing instructions that can be executed by an operating system of an optionally networked digital processing device. In other embodiments, the computer-readable storage medium is a tangible component of the digital processing device. In other embodiments, the computer-readable storage medium is optionally removable from the digital processing device. In some embodiments, the computer-readable storage medium includes, by way of non-limiting example, a CD-ROM, a DVD, a flash memory device, a solid-state memory, a disk drive, a tape drive, an optical drive, a cloud computing system and services, etc. In some cases, the program and instructions are permanently, substantially permanently, semi-permanently or non-temporarily encoded on the medium.

[0302] Computer Programs

[0303] In some embodiments, the contiguous group assembly method disclosed herein includes at least one computer program or its use. The computer program includes a sequence of instructions that are written to perform specified tasks and can be executed in the CPU of a digital processing device. Computer-readable instructions can be implemented as program modules that perform specific tasks or implement specific abstract data types, such as functions, objects, application programming interfaces (APIs), data structures, etc. Based on the disclosure provided herein, those skilled in the art will recognize that computer programs can be written in multiple versions in multiple languages.

[0304] The functions of computer readable instructions can be combined or distributed in a variety of environments as needed. In some embodiments, the computer program comprises an instruction sequence. In some embodiments, the computer program comprises multiple instruction sequences. In some embodiments, the computer program is provided from one location. In other embodiments, the computer program is provided from multiple locations. In various embodiments, the computer program comprises one or more software modules. In various embodiments, the computer program partially or entirely includes one or more network applications, one or more mobile applications, one or more stand-alone applications, one or more network browser plug-ins, extensions, add-ins or additional components or a combination thereof.

[0305] Web Application

[0306] In some embodiments, the computer program implementing the contig assembly method includes a network application. Based on the disclosure provided herein, those skilled in the art will recognize that in various embodiments, the network application utilizes one or more software frameworks and one or more database systems. In some embodiments, in a software framework (such as .NET or Ruby on Rails (RoR)). In some embodiments, the web application utilizes one or more database systems, including, by way of non-limiting example, relational, non-relational, object-oriented, associative, and XML database systems. In other embodiments, suitable relational database systems include, by way of non-limiting example, SQL Server, mySQL TM and Those skilled in the art will also recognize that in various embodiments, the web application is written in one or more versions of one or more languages. The web application may be written in one or more markup languages, presentation definition languages, client-side scripting languages, server-side coding languages, database query languages, or combinations thereof. In some embodiments, the web application is written to a certain extent in a markup language, such as Hypertext Markup Language (HTML), Extensible Hypertext Markup Language (XHTML), or Extensible Markup Language (XML). In some embodiments, the web application is written to a certain extent in a presentation definition language, such as Cascading Style Sheets (CSS). In some embodiments, the web application is written to a certain extent in a client-side scripting language, such as Asynchronous Javascript and XML (AJAX), Actionscript, Javascript, or In some embodiments, the web application is written in a server-side coding language, such as Active Server Pages (ASP), Perl, Java TM , JavaServer Pages (JSP), Hypertext Preprocessor (PHP), Python TM , Ruby, Tcl, Smalltalk, In some embodiments, the web application is written in a database query language, such as Structured Query Language (SQL). In some embodiments, the web application is integrated with an enterprise server product, such as Lotus In some embodiments, the network application includes a media player element. In a variety of other embodiments, the media player element utilizes a number of suitable multimedia technologies (including, for non-limiting examples, HTML 5, Java TM and )

[0307] Mobile Apps

[0308] In some embodiments, the computer program implementing the stack assembly method disclosed herein includes a mobile application provided to a mobile digital processing device. In some embodiments, the mobile application is provided to the mobile digital processing device at the time of manufacture. In other embodiments, the mobile application is provided to the mobile digital processing device via a computer network as described herein.

[0309] In view of the disclosure provided herein, mobile applications can be created using techniques known to those skilled in the art, using hardware, languages, and development environments known in the art. Those skilled in the art will recognize that mobile applications are written in a variety of languages. Suitable programming languages ​​include, by way of non-limiting example, C, C++, C#, Objective-C, Java, and the like. TM ,Javascript,Pascal,Object Pascal,Python TM , Ruby, VB.NET, WML, and XHTML / HTML with or without CSS or a combination thereof.

[0310] Suitable mobile application development environments are available from a variety of sources. Commercially available development environments include, by way of non-limiting example, AirplaySDK, alcheMo, Celsius, Bedrock, Flash Lite, .NET Compact Framework, Rhomobile, and WorkLight Mobile Platform. Other development environments are available for free, including, by way of non-limiting example, Lazarus, MobiFlex, MoSync, and Phonegap. In addition, mobile device manufacturers release software development kits, including, by way of non-limiting example, the iPhone and iPad (iOS) SDK, the Android TM SDK, SDK, BREW SDK, OS SDK, Symbian SDK, webOS SDK and MobileSDK.

[0311] Those skilled in the art will recognize that a number of commercial forums are available for the distribution of mobile applications, including, by way of non-limiting example, App Store, Android TM Market, App World, App Store for Palm devices, App Catalog for webOS, Marketplace, for Ovi Store on your device, Apps and DSi Shop.

[0312] Standalone Application

[0313] In some embodiments, the computer program implementing the contiguous assembly method disclosed herein comprises a stand-alone application, which is a program that runs as an independent computer process, rather than an addition to an existing process (e.g., not a plug-in). Those skilled in the art will recognize that stand-alone applications are often compiled. A compiler is a computer program that converts source code written in a programming language into binary object code (such as assembly language or machine code). Suitable compiled programming languages ​​include, by way of non-limiting example, C, C++, Objective-C, COBOL, Delphi, Eiffel, Java, and the like. TM , Lisp, Python TM , Visual Basic and VB.NET or a combination thereof. Compilation is typically performed at least in part to create an executable program. In some embodiments, the computer program includes one or more executable compiled applications.

[0314] Web browser plug-ins

[0315] In some embodiments, the stack assembly method includes a web browser plug-in. In computing, a plug-in is one or more software components that add specific functionality to a larger software application. Manufacturers of software applications support plug-ins to enable third-party developers to create capabilities that extend the application to support easy addition of new features and reduce the size of the application. When supported, the plug-in can implement the functionality of a custom software application. For example, plug-ins are commonly used in web browsers to play videos, generate interactions, scan for viruses, and display specific file types. Those skilled in the art will be familiar with a variety of web browser plug-ins, including Player, and In some embodiments, the toolbar comprises one or more web browser extensions, add-ons or add-ons. In some embodiments, the toolbar comprises one or more browser bars, tool belts or desktop belts.

[0316] In view of the disclosure provided herein, those skilled in the art will recognize that a variety of plug-in frameworks are available to enable plug-in development in a variety of programming languages, including, by way of non-limiting example, C++, Delphi, Java, and the like. TM , PHP, Python TM and VB.NET or a combination thereof.

[0317] A web browser (also known as an Internet browser) is a software application designed for use with a networked digital processing device for retrieving, presenting, and traversing information resources on the World Wide Web. Suitable web browsers include, by way of non-limiting example, Internet Chrome, Opera and KDE Konqueror. In some embodiments, the web browser is a mobile web browser. Mobile web browsers (also known as micro browsers, mini browsers, and wireless browsers) are designed for use on mobile digital processing devices, including, by way of non-limiting example, handheld computers, tablet computers, netbook computers, small notebook computers, smart phones, music players, personal digital assistants (PDAs), and handheld video game systems. Suitable mobile web browsers include, by way of non-limiting example, Browser, RIM Browser, Blazer, Browsers, for mobile devices Internet Mobile, Basic Web, Browser, Opera Mobile and PSP TM Browser.

[0318] Software Modules

[0319] In some embodiments, the contiguous group assembly method disclosed herein includes software, server and / or database modules or their use. In view of the disclosure provided herein, by techniques known to those skilled in the art, machines, software and languages ​​known in the art are used to create software modules. The software modules disclosed herein are implemented in a variety of ways. In various embodiments, the software module includes a file, a code segment, a programming object, a programming structure or a combination thereof. In other various embodiments, the software module includes multiple files, multiple code segments, multiple programming objects, multiple programming structures or a combination thereof. In various embodiments, one or more software modules include network applications, mobile applications and stand-alone applications for non-limiting examples. In some embodiments, the software module is in a computer program or application. In other embodiments, the software module is in more than one computer program or application. In some embodiments, the software module is hosted on a machine. In other embodiments, the software module is hosted on more than one machine. In other embodiments, the software module is hosted on a cloud computing platform. In some embodiments, the software module is hosted on one or more machines located at one location. In other embodiments, the software module is hosted on one or more machines located at more than one location.

[0320] database

[0321] In some embodiments, the contiguous group assembly method disclosed herein includes one or more databases or their use. In view of the disclosure provided herein, those skilled in the art will recognize that many databases are suitable for storage and retrieval of contiguous group information. In various embodiments, suitable databases include, by way of non-limiting example, relational databases, non-relational databases, object-oriented databases, object databases, entity relationship model databases, associative databases, and XML databases. In some embodiments, the database is based on the Internet. In other embodiments, the database is network-based. In other embodiments, the database is based on cloud computing. In other embodiments, the database is based on one or more local computer storage devices.

[0322] Diagnostic Applications

[0323] The systems and methods herein are applicable to the selection or evaluation of drugs or other treatment regimens. Through the practice of the disclosure herein, tissues such as cancer tissues are evaluated to determine structural rearrangements that indicate candidate drugs. For example, local density variation or local density variation patterns indicate changes in specific one or more genes in some cases. For example, the rearrangement involved in the analysis may involve gene truncation, deletion or fusion, thereby forming a genomic background that is known or suspected to respond to a specific therapy. An analysis indicating a treatment strategy is performed, and a drug is indicated. Typically, a drug or other treatment regimen is recommended to a medical professional or patient, or applied to a patient to address a medical condition associated with the analyzed sample.

[0324] Alternatively or in combination, the systems and methods disclosed herein are used to monitor the success of a drug or other treatment regimen applied to an individual, such as an individual whose disease being treated involves a genomic rearrangement. Samples are obtained and analyzed as disclosed herein to identify local density patterns. Typically but certainly, local density variations involve specific genomic rearrangements associated with the disease, suggesting treatment methods, or indicating disease progression (such as by the abundance of rearrangements in the sample). A treatment regimen such as drug therapy is performed (alone or in combination with other treatment steps or other steps not involving drugs) to treat or alleviate the symptoms of the condition. A second sample is obtained and analyzed as disclosed herein to identify local density patterns. The pattern or resulting analysis is compared with a pattern or analysis observed before or early in the treatment regimen to evaluate the efficacy of the regimen, such as the efficacy of a drug in reducing the abundance of a specific rearrangement in a tumor, or the efficacy of a surgical intervention or other treatment regimen in removing or reducing tissue suspected of being the cause of a specific tissue disease (such as a cancer tumor) or associated with a specific tissue disease. Evaluation variously includes stopping the treatment regimen, reducing the treatment regimen, initiating a second treatment regimen, maintaining the treatment regimen unchanged, increasing the treatment regimen, substituting monitoring for treatment or other regimen inputs.

[0325] Numbered embodiments related to the present disclosure

[0326] The present disclosure is further illustrated by reference to the following numbered embodiments, which are given in numerical order, but should be understood to be readily interrelated to each other and to the remainder of the specification in addition to the interrelationship indicated by the following numbers. The numbered embodiments are given to further illustrate the disclosure herein and to support claims that reference the subject matter of the embodiments. 1. A method for nucleic acid structural variant detection, comprising: a) mapping read pair information onto a reference nucleic acid scaffold; b) assigning read pair positions to a first bin such that the read pair midpoints fall within the first bin nucleic acid position range and the read pair intervals fall within the first bin interval range; and c) estimating copy number variation based on the mappability values ​​of the first bin. 2. The method according to embodiment 1, further comprising normalizing the copy number variation. 3. The method according to embodiment 1, further comprising displaying mappability by plotting the mapped read densities of two samples against each other. 4. A method for nucleic acid structural variant detection, comprising: a) mapping read pair information onto a reference nucleic acid scaffold; b) assigning read pair positions to a first bin such that the read pair midpoints fall within the first bin nucleic acid position range and the read pair intervals fall within the first bin interval range; c) generating a two-dimensional image of the read pair information; wherein each pixel represents a bin; d) calculating a z-score for at least one set of four pixels in the image that share a common corner; wherein the z-score is represented by the contrast between adjacent pixels; and e) identifying a candidate hit when the z-score exceeds a threshold. 5. The method according to any one of embodiments 1-4, wherein the reference nucleic acid scaffold is a genome. 6. The method according to any one of embodiments 1-4, wherein each data set is obtained from a different paired-end read direction. 7. The method according to any one of embodiments 1-4, wherein the candidate hit is a translocation. 8. The method according to any one of embodiments 1-4, wherein the candidate hit is an inversion. 9. The method according to any one of embodiments 1-4, wherein the candidate hit is a deletion. 10. The method according to any one of embodiments 1-4, wherein the candidate hit is a duplication. 11. A method according to any one of embodiments 1-4, wherein the candidate hit is an inter-chromosomal structural variation. 12. A system for modeling a mixture of allele variation in a sample, comprising: a collection of weighted genomic scaffold models, wherein each genomic scaffold model includes a collection of weighted chromosomes, wherein each chromosome is a linear graph of bins in the genomic scaffold; and a module for calculating the log-likelihood ratio of at least two genomic scaffold models to predict whether a read pair sampled from a library will fall into a bin. 13. A system according to any one of embodiments 1-12, further comprising at least one feature detector module, wherein the at least one feature detector module proposes candidate modifications to the genomic scaffold model.14. The system of any one of embodiments 1-13, wherein the at least one feature detector module determines bin boundaries for sequence variants. 15. The system of any one of embodiments 1-14, wherein the sequence variant is a translocation. 16. The system of any one of embodiments 1-14, wherein the sequence variant is an inversion. 17. The system of any one of embodiments 1-14, wherein the sequence variant is a deletion. 18. The system of any one of embodiments 1-14, wherein the sequence variant is a duplication. 19. The system of any one of embodiments 1-12, further comprising a module for generating a substitution model based on input from the at least one feature detector module. 20. A method for modeling allelic variation in a sample, comprising: a) generating a collection of weighted genomic scaffold models, wherein each genomic scaffold model comprises a collection of weighted chromosomes, wherein each chromosome is a linear map of bins in the genomic scaffold; b) calculating a score based on the ability of the model to describe sequencing information of read pairs mapped to a reference sequence, wherein a higher score value indicates a more predictive model; and c) iteratively adding additional models to maximize the score value. 21. The method according to any one of embodiments 1-20, wherein the read pair sequencing information includes an inversion. 22. The method according to any one of embodiments 1-20, wherein the read pair sequencing information includes a translocation. 23. The method according to any one of embodiments 1-20, wherein the read pair sequencing information includes a duplication. 24. The method according to any one of embodiments 1-20, wherein the read pair sequencing information includes a deletion. 25. The method according to any one of embodiments 1-21, further comprising detecting a feature, wherein detecting a feature comprises joining or separating bins in the model to increase the score value. 26. The method according to any one of embodiments 1-20, wherein the sample is a cancer cell. 27. A method for nucleic acid structural variant detection, comprising: a) mapping read pair information to a predicted nucleic acid scaffold; b) assigning a read pair position to a first bin such that a read pair midpoint falls within a first bin nucleic acid position range and a read pair interval falls within a first bin interval range; c) generating a two-dimensional image of the read pair information; wherein each pixel represents a bin; and d) identifying at least one feature in the two-dimensional image corresponding to two sequence fragments connected by a common connecting sequence fragment. 28. The method of any one of embodiments 1-27, comprising assembling the two sequence fragments connected by the common connecting sequence fragment in the correct order. 29. The method of any one of embodiments 1-27, wherein the method comprises discarding features corresponding to false positives. 30. A method comprising: mapping read pair sequence information onto a sequence scaffold; and identifying local variations in the density of a plurality of read pair symbols so mapped.31. The method according to any one of embodiments 1-30, comprising assigning local variations of the density to corresponding structural arrangement features. 32. The method according to any one of embodiments 1-30, comprising reconstructing the sequence scaffold so as to reduce local variations of the density. 33. The method according to any one of embodiments 1-30, wherein mapping the read pair sequence information onto the sequence scaffold comprises positioning a symbol indicating a read pair so that the distance of the symbol from the axis representing the sequence scaffold indicates the distance from the mapping position of the first read of the read pair on the sequence scaffold to the mapping position of the second read of the read pair on the sequence scaffold, and so that the position of the symbol relative to the axis representing the sequence scaffold indicates the average of the mapping position of the first read of the read pair and the mapping position of the second read of the read pair. 34. The method according to any one of embodiments 1-31, wherein reconstructing the sequence scaffold comprises reordering at least some contigs of the sequence scaffold. 35. The method according to any one of embodiments 1-31, wherein reconstructing the sequence scaffold comprises reorienting at least one contig of the sequence scaffold. 36. The method according to any one of embodiments 1-31, wherein reconstructing the sequence scaffold comprises introducing a break into at least one contig of the sequence scaffold. 37. The method according to any one of embodiments 1-36, further comprising introducing a sequence present at one edge of the break onto a second edge of the break. 38. The method according to any one of embodiments 1-30, wherein reconstructing the sequence scaffold comprises translocating a segment of a first contig into an internal region of a second contig. 39. The method according to any one of embodiments 1-30, wherein mapping read pair sequence information onto a sequence scaffold comprises assigning read pair information to a plurality of bins. 40. The method according to any one of embodiments 1-30, wherein identifying local variations in density comprises identifying regions of local low density with a symbol. 41. The method according to any one of embodiments 1-30, wherein identifying local variations in density comprises identifying regions of local high density with a symbol. 42. The method according to any one of embodiments 1-30, wherein identifying local variations in density comprises identifying a density at a first position and a density at a second position, wherein the density at the first position and the density at the second position are significantly different. 43. The method of any one of embodiments 1-42, wherein the first position and the second position are adjacent. 44. The method of any one of embodiments 1-42, wherein the first position and the second position are equidistant from the sequence scaffold. 45. The method of any one of embodiments 1-30, wherein identifying local variation in density comprises obtaining an expected density at a first position and an observed density at the first position.46. ​​A method according to any one of embodiments 1-45, wherein the expected density at the first position is a density predicted by a density gradient that decreases monotonically with increasing distance from the axis representing the sequence scaffold. 47. A method according to any one of embodiments 1-30, wherein a local density variation equal to a fraction of an integer value of the ploidy of the sample indicates an event of sample ploidy complementation in that proportion. 48. A method according to any one of embodiments 1-30, wherein the scaffold represents a cancer cell genome. 49. A method according to any one of embodiments 1-30, wherein the scaffold represents a transgenic cell genome. 50. A method according to any one of embodiments 1-30, wherein the scaffold represents a gene-edited genome. 51. A method according to any one of embodiments 1-32, wherein the scaffold has an N50 increased by at least 20% after the reconstruction. 52. A method comprising: obtaining a scaffold comprising sequence scaffold information; obtaining paired read information; deploying the paired read information so that at least some read pair information is depicted to indicate the position of each read in the read pair relative to the scaffold and to indicate the distance of one read to another read mapped on the scaffold; and identifying local variation in the density of the deployed paired read information. 53. The method according to any one of embodiments 1-52, comprising assigning local variations of the density to corresponding structural arrangement features. 54. The method according to any one of embodiments 1-52, comprising reconfiguring the scaffold to reduce the local variations. 55. The method according to any one of embodiments 1-52, wherein obtaining a scaffold comprising sequence scaffold information comprises sequencing a nucleic acid sample. 56. The method according to any one of embodiments 1-52, wherein obtaining a scaffold comprising sequence scaffold information comprises receiving digital information representing a nucleic acid sample. 57. The method according to any one of embodiments 1-52, comprising obtaining a predicted density distribution for the deployed read pair information. 58. The method according to any one of embodiments 1-57, wherein the identification comprises identifying a significant difference between the predicted density distribution and the depicted read pair information density. 59. The method according to any one of embodiments 1-52, wherein identifying local variations comprises identifying a density perturbation having a density peak at the vertex of a right angle. 60. The method according to any one of embodiments 1-59, wherein the vertex of the right angle points to an axis representing the scaffold. 61. The method of any one of embodiments 1-52, wherein obtaining paired-end read information comprises cross-linking unextracted nucleic acids. 62. The method of any one of embodiments 1-52, wherein obtaining paired-end read information comprises cross-linking nucleic acids bound in chromatin. 63. The method of any one of embodiments 1-62, wherein the chromatin is native chromatin.64. The method according to any one of embodiments 1-52, wherein obtaining paired-end read information comprises binding a nucleic acid to a nucleic acid binding moiety. 65. The method according to any one of embodiments 1-52, wherein obtaining paired-end read information comprises generating reconstructed chromatin. 66. The method according to any one of embodiments 1-52, wherein deploying the paired read information comprises assigning read pair information to a plurality of bins. 67. The method according to any one of embodiments 1-52, wherein reconstructing the sequence scaffold comprises reordering at least some contigs of the sequence scaffold. 68. The method according to any one of embodiments 1-54, wherein reconstructing the sequence scaffold comprises reorienting at least one contig of the sequence scaffold. 69. The method according to any one of embodiments 1-54, wherein reconstructing the sequence scaffold comprises introducing a break into at least one contig of the sequence scaffold. 70. The method according to any one of embodiments 1-69, further comprising introducing a sequence at one edge of the break onto a second edge of the break. 71. The method according to any one of embodiments 1-54, wherein reconstructing the sequence scaffold comprises translocating a segment of a first contig into an interior region of a second contig. 72. The method of any one of embodiments 1-52, wherein the scaffold represents a cancer cell genome. 73. The method of any one of embodiments 1-52, wherein the scaffold represents a transgenic cell genome. 74. The method of any one of embodiments 1-52, wherein the scaffold represents a gene-edited genome. 75. The method of any one of embodiments 1-52, wherein the scaffold has an N50 that is increased by at least 20% after the remodeling. 76. The method of any one of embodiments 1-52, wherein a local density variation that is equal to a fraction of an integer value of the ploidy of the sample indicates an event of sample ploidy complementation in that fraction. 77. A method of identifying a structural rearrangement in a sample relative to a sequence scaffold, comprising: mapping read pair sequence information onto a sequence scaffold; identifying a local density variation having a right angled side pointing to an axis corresponding to the sequence scaffold and having bilateral symmetry along a line bisecting the right angled side; and classifying the sample as having a simple translocation relative to the sequence scaffold, the sequence scaffold comprising a segment length at least as long as the longest furthest mapped read of the local density variation from the translocation site. 78. A method for identifying structural rearrangements in a sample, comprising: mapping read pair sequence information onto a sequence scaffold; identifying local density variations, the local density variations having a right-angled side pointing to an axis corresponding to the sequence scaffold; identifying sub-regions of the local density variations, the sub-regions destroying bilateral symmetry along a line bisecting the right-angled side; and classifying the sample as having a translocation relative to the sequence scaffold, the sequence scaffold comprising a segment of the sequence to which a symmetry-restoring read pair population is mapped.79. A method for identifying structural rearrangements relative to a sequence scaffold in a sample, comprising: mapping read pair sequence information onto a sequence scaffold; identifying local density variations having a right-angled side pointing to an axis corresponding to the sequence scaffold; obtaining an expected read pair density distribution curve; and identifying a scaffold segment to which a read pair containing the local density variation is mapped; repositioning the scaffold segment so that the read pair containing the local density variation is mapped to an area indicated by the expected read pair density distribution curve to obtain a density of the local density variation. 80. A computer monitor configured to display the results of a method according to any one of embodiments 1-79. 81. A computer system configured to perform the computational steps of a method according to any one of embodiments 1-79. 82. A visual representation of mapped read pair data according to any one of embodiments 1-79. 83. A method for nucleic acid structural variant detection, comprising: mapping read pair information onto a predicted nucleic acid scaffold; obtaining a structural variant hypothesis; calculating a likelihood parameter for the structural variant hypothesis to fit the read pair information; and classifying the nucleic acid sample as having the structural variant hypothesis if the likelihood parameter for the hypothesis is greater than a second likelihood parameter for a second hypothesis, wherein mapping the read pair information onto the predicted nucleic acid scaffold comprises assigning a read pair position to a read pair such that the read pair is assigned to its midpoint on the predicted nucleic acid scaffold on one axis; and such that the read pair is assigned a value corresponding to its read pair spacing on a second axis. 84. The method of any one of embodiments 1-83, wherein the read pair comprises a first segment mapped to a first region of a nucleic acid molecule and a second segment mapped to a second region of the nucleic acid molecule, the first segment and the second segment being non-adjacent and sharing a common phase. 85. The method of any one of embodiments 1-83, wherein the read pair position is assigned to a first bin if the read pair midpoint falls within a first bin nucleic acid position range and the read pair spacing falls within a first bin spacing range. 86. The method of any one of embodiments 1-85, wherein the first bin nucleic acid position range is a regular spacing of the predicted nucleic acid scaffold. 87. The method of any one of embodiments 1-85, wherein the first bin spacing range is a logarithmic spacing of the full spacing range of the read pair information. 88. The method of any one of embodiments 1-85, wherein the first bin nucleic acid range is a regular spacing of the nucleic acid scaffold, and wherein the first bin spacing range is a logarithmic spacing of the full spacing range of the read pair information. 89. The method of any one of embodiments 85-88, wherein the read pair position is assigned to a second bin if a read pair midpoint falls within a second bin nucleic acid position range and a read pair spacing falls within a second bin spacing range.90. The method of any one of embodiments 1-89, wherein substantially all of the read information is binned. 91. The method of any one of embodiments 85-90, wherein calculating the likelihood parameter comprises determining a likelihood contribution of the first bin. 92. The method of any one of embodiments 1-91, wherein the likelihood contribution of the first bin comprises a first likelihood factor proportional to the count of the read pairs mapped to the first bin. 93. The method of any one of embodiments 1-91, wherein the likelihood contribution of the first bin comprises a second likelihood factor proportional to an area of ​​the first bin. 94. The method of any one of embodiments 1-91, wherein the likelihood contribution of the first bin comprises a first likelihood factor proportional to the count of the read pairs mapped to the first bin, and wherein the likelihood contribution of the first bin comprises a second likelihood factor proportional to an area of ​​the first bin. 95. The method of any one of embodiments 1-94, comprising determining a likelihood contribution of a second bin that does not overlap in area with the first bin. 96. The method of any one of embodiments 1-95, wherein the likelihood parameter comprises a likelihood contribution of the first bin and a likelihood contribution of the second bin. 97. The method of any one of embodiments 1-96, wherein the likelihood parameter comprises a likelihood contribution of a third bin. 98. The method of any one of embodiments 1-97, wherein the likelihood parameter comprises a likelihood contribution of the reads to the information for substantially all bins. 99. The method of any one of embodiments 78-98, wherein the hypothesis comprises a structural variation having a left margin and a length. 100. The method of any one of embodiments 1-99, wherein the structural variation has an orientation of at least one of a deletion, an inversion, a direct repeat, an outward inverted repeat, and an inward inverted repeat. 101. The method of any one of embodiments 99-100, wherein the second hypothesis comprises a structural variant that differs in at least one of a left margin, a length, and a structural orientation. 102. The method of any one of embodiments 1-101, wherein the nucleic acid structural variant is homozygous in the nucleic acid sample. 103. The method of any one of embodiments 78-101, wherein the nucleic acid structural variant is heterozygous in the nucleic acid sample. 104. A method of displaying putative structural variation in a nucleic acid sample, comprising the steps of assigning a population of sequence reads to a population of numbered bins, and assigning a likelihood parameter for reads comprising an edge of a structural variation that falls within a first bin of the population of bins, wherein the likelihood parameter for the first bin comprises a first likelihood component comprising the number of reads mapped to the first bin, and a second component comprising the area of ​​the first bin.105. The method of any one of embodiments 1-104, comprising plotting the likelihood of a structural variation as a function of the number of bins. 106. The method of any one of embodiments 1-104, wherein the likelihood parameter for the first bin comprises the convolution of a first likelihood component and a second component, the first likelihood component comprising the number of reads mapped to the first bin, the second component comprising the area of ​​the first bin. 107. The method of any one of embodiments 1-106, wherein the likelihood parameter comprises a likelihood component associating a structural variant prediction with the number of reads mapped to the first bin and a likelihood component comprising the area of ​​the first bin. 108. The method of any one of embodiments 1-104, wherein the population of bins shares a common bin width across a fixed nucleic acid distance. 109. The method of any one of embodiments 1-104, wherein there is variation in bin height between members of the population of bins. 110. The method of any one of embodiments 1-109, wherein bin height appears constant when plotted on a logarithmic axis. 111. The method of any one of embodiments 1-104, wherein the likelihood parameter relates to the probability that a sequence read comprising a junction having a structural variation with a left edge and a length maps to the first bin. 112. The method of any one of embodiments 1-111, wherein the structural variation has an orientation of at least one of a deletion, an inversion, a direct repeat, an outward inverted repeat, and an inward inverted repeat. 113. The method of any one of embodiments 1-104, wherein the sequence read comprises a read pair. 114. The method of any one of embodiments 1-113, wherein the read pair comprises a first segment mapped to a first region of a nucleic acid molecule and a second segment mapped to a second region of the nucleic acid molecule, the first segment and the second segment being non-adjacent and sharing a common phase. 115. A method for identifying structural variants in a nucleic acid sample, comprising the steps of: obtaining mapped read pair data for the nucleic acid sample; obtaining a nucleic acid scaffold sequence; obtaining likelihood probability information for each of a plurality of structural variant hypotheses comparing the read pair data to the nucleic acid scaffold sequence; and identifying the most likely hypothesis among the structural variant hypotheses; wherein the method evaluates at least 10 Mb of nucleic acid scaffold sequence per minute. 116. The method of any one of embodiments 1-115, comprising mapping the read pair information to the nucleic acid scaffold sequence; obtaining a structural variant hypothesis; calculating a likelihood parameter of the conformity of the structural variant hypothesis with the read pair information; and classifying the nucleic acid sample as having the structural variant hypothesis if the likelihood parameter of the hypothesis is greater than a second likelihood parameter of a second hypothesis.117. The method of any one of embodiments 1-116, wherein mapping read pair information onto the nucleic acid scaffold sequence comprises: assigning read pair positions to read pairs such that the read pairs are assigned to their midpoints on the predicted nucleic acid scaffold on one axis; and the read pairs are assigned values ​​corresponding to their read pair spacings on a second axis. 118. The method of any one of embodiments 116-112, wherein the read pairs comprise a first segment mapped to a first region of a nucleic acid molecule and a second segment mapped to a second region of the nucleic acid molecule, the first segment and the second segment being non-adjacent and sharing a common phase. 119. The method of any one of embodiments 1-117, wherein the read pair positions are assigned to a first bin if the read pair midpoints fall within a first bin nucleic acid position range and the read pair spacing falls within a first bin spacing range. 120. The method of any one of embodiments 1-119, wherein the first bin nucleic acid position range is a regular spacing of nucleic acid scaffolds. 121. The method of any one of embodiments 1-119, wherein the first bin spacing range is a logarithmic spacing of the complete spacing range of the read pair information. 122. The method of any one of embodiments 1-119, wherein the first bin nucleic acid position range is a regular spacing of nucleic acid scaffolds, and wherein the first bin spacing range is a logarithmic spacing of the complete spacing range of the read pair information. 123. The method of any one of embodiments 119-122, wherein the read pair position is assigned to a second bin if the read pair midpoint falls within a second bin nucleic acid position range and the read pair spacing falls within the second bin spacing range. 124. The method of any one of embodiments 1-123, wherein substantially all read information is binned. 125. The method of any one of embodiments 119-119, wherein calculating the likelihood parameter comprises determining a likelihood contribution of the first bin. 126. The method of any one of embodiments 1-125, wherein the likelihood contribution of the first bin comprises a first likelihood factor proportional to the count of the read pairs mapped to the first bin. 127. The method of any one of embodiments 1-120, wherein the likelihood contribution of the first bin comprises a second likelihood factor proportional to the area of ​​the first bin. 128. The method of any one of embodiments 1-120, wherein the likelihood contribution of the first bin comprises a first likelihood factor proportional to the count of the read pairs mapped to the first bin, and wherein the likelihood contribution of the first bin comprises a second likelihood factor proportional to the area of ​​the first bin. 129. The method of any one of embodiments 1-123, comprising determining a likelihood contribution of a second bin that does not overlap in area with the first bin.130. The method of any one of embodiments 1-124, wherein the likelihood parameter comprises a likelihood contribution of the first bin and a likelihood contribution of the second bin. 131. The method of any one of embodiments 1-130, wherein the likelihood parameter comprises a likelihood contribution of a third bin. 132. The method of any one of embodiments 1-126, wherein the likelihood parameter comprises a likelihood contribution of the reads to the information for substantially all bins. 133. The method of any one of embodiments 115-127, wherein the hypothesis comprises a structural variation having a left margin and a length. 134. The method of any one of embodiments 1-128, wherein the structural variation has an orientation of at least one of a deletion, an inversion, a direct repeat, an outward inverted repeat, and an inward inverted repeat. 135. The method of any one of embodiments 134-129, wherein the second hypothesis comprises a structural variant that differs in at least one of a left margin, a length, and a structural orientation. 136. The method of any one of embodiments 111-130, wherein the nucleic acid structural variant is homozygous in the nucleic acid sample. 137. The method of any one of embodiments 111-130, wherein the nucleic acid structural variant is heterozygous in the nucleic acid sample. 138. A method of selecting a treatment regimen, comprising: performing the method of any one of the preceding embodiments, identifying a rearrangement, and identifying a treatment regimen consistent with the rearrangement. 139. The method of any one of embodiments 1-133, wherein the treatment regimen comprises drug administration. 140. The method of any one of embodiments 1-133, wherein the treatment regimen comprises tissue resection. 141. A method of evaluating a treatment regimen, comprising: performing the method of any one of the preceding embodiments a first time, administering the treatment regimen, and performing the treatment regimen a second time. 142. The method of any one of embodiments 1-136, comprising discontinuing the treatment regimen. 143. The method of any one of embodiments 1-136, comprising increasing the dose of the treatment regimen. 144. The method of any one of embodiments 1-136, comprising decreasing the dose of the treatment regimen. 145. The method of any one of embodiments 1-136, comprising continuing the treatment regimen. 146. The method of any one of embodiments 136-140, wherein the treatment regimen comprises a drug. 147. The method of any one of embodiments 136-140, wherein the treatment regimen comprises a surgical intervention.

[0327] Discussion of the Figures

[0328] exist Figure 1An exemplary workflow of 8 steps for a method for processing paired-end read data is observed in . Exemplary steps include read mapping (mapping paired sequence reads from one individual against a reference), read binning (grouping reads by one or more properties), copy number estimation (copy number variation, CNV), normalization, de novo feature detection, breakpoint refinement, candidate scoring, and reporting. In some cases, steps are repeated or skipped entirely when analyzing paired-end read data.

[0329] exist Figure 2A-2C Plots of observed pairs in , each plot having bins corresponding to the range of midpoint positions of mapped read pairs on the x-axis, with a scale from 0 to 12,000 bases in 20,000 bp increments, and on the y-axis corresponding to the estimated copy number, with a logarithmic scale between 0.1 and 10. Figure 2A Reference sample CT407 (top), Figure 2B CT418 (top) and Figure 2C In CT416, most bases exist in a single copy, indicated by the high plot density region in the center of the vertical axis. Figure 2A The bottom picture of CT410 and Figure 2B The samples represented by CT417 in Figure 1 show significant deviations from distance 1, where the bins have more or less than one copy number. For example, sample CT410 has a copy number gain for bins at approximately 10,000 to 10,500 bases. Figure 2D A two-dimensional scatter plot is shown with the copy number of sample CT410 on the x-axis and the copy number of CT407 on the y-axis, with each point representing the copy number of the corresponding bin in each sample. Most of the points are concentrated at the coordinate (1,1) on the y=x diagonal, which corresponds to a single copy at that bin in both samples. Points that do not fall near the diagonal indicate a significant difference in copy number between the two samples. For example, the bin corresponding to (100,10) has a 10-fold increase in copy number in CT410 compared to CT407.

[0330] exist Figure 3A Observed in the figure for chromosome 7 of sample NA12878, the midpoint position of the mapped read pairs is on the x-axis and the scale is 5.31×10 7 to 5.36×10 7 base pairs in increments of 0.01×10 7 , and the read pair intervals are plotted on the y-axis, with a scale of 0 to 200,000 bases (in increments of 20,000 bases). The plot does not show any obvious structural variation, as evidenced by the fact that most points fall near 0 on the y-axis. This suggests that most read pairs correspond to adjacent segments on the scaffold. Figure 3B and Figure 3C It shows that 5.41×10 7 to 5.46×10 7 The x-axis scale is 0 to 200,000 (in increments of 20,000 bases) and the y-axis scale is 100 to 100,000 (logarithmic scale). In these figures, it is observed that the 7 and 5.44×10 7 There are inversions between bases, where there are gaps in the data. Figure 3D In the example, an exemplary depiction of an inversion between positions a and b is observed, wherein approximately half of the points (grey) remain near the axis, and the other half are reflected above the midpoint between positions a and b. In this example, the light-colored points that remain near the axis indicate heterozygous inversions, where only one chromosome in a pair is inverted. In some cases, the figure is rotated 45 degrees, with the x-axis located on the y=-x diagonal.

[0331] exist Figure 4A Examples of various structural variations observed in Figure 1, which manifest as the redistribution of mapped read pairs into the region formed by lines at a 45 degree angle to the x-axis. Figure 4B Depicted is a numerical system for defining the density region formed by lines that make an angle of 45 degrees to the axis. Figure 4C-Figure 4G Depicted are exemplary methods for defining density regions for various structural variations. In some cases, the density regions create kernel patterns. The defined patterns are variously used to predict density variations that indicate deviations between mapped read pair data and scaffolds. For example, in some cases, Figure 4C , Figure 4D , Figure 4E , Figure 4F and Figure 4G The local density variation regions of expected deletions, inversions, direct tandem repeats, inverted tandem repeats (right) or inverted tandem repeats (left) are defined respectively. Exemplary equations for defining the predicted variation in density for each region in regions 0-3 are shown on the left side of each figure.

[0332] exist Figure 5A A graph of predicted structural variations observed in , the graph comprising an x-axis representing the number of 200 base pair bins, scaled from 0 to 80,000 in intervals of 10,000, and a y-axis representing the log likelihood ratio (LLR), scaled from -250 to 150 in intervals of 50. In some cases, the log likelihood ratio represents the ratio of the likelihood that the structural variation has occurred to the likelihood that the structural variation has not occurred. Higher values ​​indicate a more likely variation, for example, the spike observed at approximately bin 36,000 corresponds to a known inversion. Figure 5BA graph of predicted structural variations observed in the , the graph includes an x-axis of the number of 200 base pair bins, scaled from 0 to 80,000, with intervals of 10,000, and a y-axis representing the log likelihood ratio (LLR), scaled from -120 to 40, with intervals of 20. In this example, the relative negative values ​​between bins of about 55000 to 68000 indicate the presence of a heterozygous deletion of 10 Kb. Figure 5C , the figure includes an x-axis representing the number of 200 base pair bins, scaled from 0 to 80,000, with intervals of 10,000, and a y-axis representing the log likelihood ratio (LLR), scaled from -100 to 60, with intervals of 20. In this example, the relatively negative values ​​between bins of approximately 55000 to 68000 indicate the presence of a 26 Kb heterozygous duplication (L).

[0333] exist Fig. 6A and Figure 6B In , an exemplary read distribution pattern is observed that in some cases depicts a reciprocal translocation, in this case a square, divided into four regions. In some cases, the pattern is a kernel or feature. In this case, the read density is distributed in the diagonal region formed by the intersection of two lines. In Figure 6C In , regions depicted as foreground (fg) and background (bg) regions are observed and compared as the ratio of fg to bg to establish a z-score in some cases. The z-score is often used to identify features from noise. Fig.6D In the figure, a plot of read pair data mapped on a scaffold is observed, where features are identified (circled). In some cases, areas of higher or lower read density are not reflected in the center of the square (upper right circle), in contrast, the lower left feature shows density reflected in the center of the square. In this example, the read pair density decreases away from the center of the square (where the highest density is found) in a gradient at a 45 degree angle. In some cases, the "bow tie" structure represented by the two circled features in the lower left corresponds to a translocation.

[0334] exist Figure 7 In the image, read pairs mapped to the scaffolds are observed, illustrating intrachromosomal rearrangements, as shown by regions of abnormally high or low read density that are offset from the diagonal y=-x axis. These regions located off the diagonal axis correspond to mapped read pairs that are separated by a distance longer than the reads, indicating potential biases in the scaffold assembly.

[0335] exist Fig. 8AIn , an illustration of a “secondary junction” assembly situation is observed, where two different assembly outcomes are possible by analyzing only the primary read pairs. The three sequences above the arrows in each group correspond to the native sequence arrangement (scaffolds): sequences ab, cde, and fg. However, rearrangements of segments in the sequences (indicated by arrows) result in two potential arrangements: ade and cdg, or adg, which are indistinguishable by primary read pair analysis, as both potential rearrangements will result in read pairs of the rearranged sequences mapping segment a to d, and d to g. In Figure 8B In , a graphic depicting read pair data mapped to a scaffold is observed, where the data on the axis is not shown. Two features were identified (the shaded box represents the read pair density, in which the intensity decreases at right angles along a gradient extending away from the diagonal axis, marked with symbols of smaller and larger circles touching each other). The alphabetical linear arrangement of fragments ag serves as a scaffold. The read pair data from the two "off-axis" features indicates a connection between fragments ad and dg. In addition, the lack of a signal marked by the concentric circles indicates that fragments a and g are not connected by the intervening sequence d. In Figure 8C A similar plot is observed, depicting the expected pattern of ADG connectivity. The connectivity of AD and DG is illustrated by the features identified at the small and large circle symbols. Although fragments A and G are not directly connected, the shaded region of the read pair corresponding to the bridging intervening sequence D is observed, and the features corresponding to AF and CG are absent (concentric circles), further supporting the hypothesis of ADG connectivity. Fig.8D In , similar plots are observed, depicting the expected pattern of adg junctions, with key features visible in the shaded boxes. In some cases, a "bridge" feature corresponding to ag indicates a false-positive fusion call between fragments a and g. In other cases, a feature at dg indicates a false-positive fusion call where there are no other fragments to the left of fragment d in dg. Fig. 8EAt , a graph showing how the abundance (g) and gap size / distance (γ) of the read pairs in the mixture predicts the expected change in density (contour line) is observed. For example, the left figure depicts that the read density (from the middle of the contour line) decreases rapidly when the distance between the read pairs (g) is small and the abundance is low. The right figure depicts that the read density (from the middle of the contour line) decreases rapidly when the distance between the read pairs (g) is large and the abundance is high. In some cases, the rate at which the read density decreases is used to predict the blocking edge between sequence fragments. For example, a sharp and rapid decrease in the read density adjacent to a kernel indicates a lack of adjacent kernels. In some cases, a comparison of the expected read density of a region is used to minimize false positive kernel determinations. Typically, the putative kernel will have an expected read density higher than the terminal fragment (connected only to an additional fragment), and will therefore not be identified as a terminal fragment. Alternatively, the putative kernel will have an expected read density less than the fusion event, and will therefore not be identified as a fusion event. In some cases, the rapid decrease in density is referred to as a "staircase", which contrasts with the gradual change in density. The expected density may also be defined or described by geometric considerations such as symmetry. For example, symmetrical changes in read density indicate isolated deviations from the scaffold model, wherein asymmetrical changes in read density optionally indicate the presence of additional adjacent deviations.

[0336] exist Fig. 9 In Figure 1, an image of read pairs from two genes mapped to a scaffold is observed, illustrating structural variation, as shown by regions of abnormally high or low read density that deviate from the diagonal y=-x axis. The bow-tie-shaped density distribution in the upper right and lower left boxed regions indicates a reciprocal translocation between the genes ETV6 and NTRK3.

[0337] exist Figure 10A-10C In Figure 2, the results of image-based analysis are observed for the same chromosome pair compared in three different samples. The circled regions correspond to identified features representing structural variation.

[0338] exist Figures 11A-11C In the , it was observed that chromosome 1 and chromosome 7 ( Fig.11A ), chromosome 2 pair, chromosome 5 ( Fig. 11B ) and chromosome 1 to chromosome 1 ( Fig. 11C ) of the normalized median read density (over 10 samples).

[0339] exist Fig. 12A and Fig. 12B In Figure 1, images depicting various binning methods for mapped read pair data are observed, which place the read pairs into groups. Fig. 12A shows equal bin sizes, while Fig. 12B Bin cell interpolation is shown.

[0340] exist Fig.13 In Figure 1, an image depicting the whole genome scan analysis pipeline is observed, where the identified features correspond to structural variants. Sample calls made by the analysis pipeline are shown as white circles. Fig.13 A map of chromosome 3 to chromosome 6 is shown, with 250k bins.

[0341] exist Fig.14A In FIG. 1 , the probability of inserts within a particular range is observed versus the insertion distance in base pairs (bp) for preserved samples (e.g., FFPE samples) analyzed by the techniques of the present disclosure. Fig. 14B Similar plots are observed for samples analyzed using the Chicago method in Figure 2. In both plots, the x-axis shows the insertion distance (bp), from 0 to 300,000 (in 50,000 bp increments), while the y-axis shows the probability of an insertion at that distance, starting at 10 at the top of the axis. 0 To the bottom of the axis 10 -8 (Logarithm).

[0342] exist Fig.15A and Fig. 15B In Figure 1, a map of the mapping positions of read pairs on a reference sequence such as GRCh38 is observed. These read pairs were generated by proximity ligation of DNA from reassembled chromatin and are plotted near structural differences between GM12878 and the reference. Fig.15A In the , the x-axis is read position 1 (in Mb) and the scale is 54.2 to 54.55 in 0.05Mb increments. The y-axis is read position 2 (in Mb) and the scale is 54.15 to 54.55 in 0.05Mb increments. Fig. 15B , the x-axis is read position 1 (in Mb) and the scale is 78.85 to 79.15 in increments of 0.05 Mb. The y-axis is read position 2 (in Mb) and the scale is 78.8 to 79.2 in increments of 0.05 Mb. Each read pair generated is represented above and below the diagonal. Above the diagonal, the shading represents the map quality score on the scale shown; below the diagonal, the shading represents the inferred unit type phase of the read pair generated based on the overlap with the phased SNP. In some embodiments, the generated map depicts an inversion with a flanking repeat region, such as Fig. 15B In some embodiments, the generated graph depicts the data of the phased heterozygous deletion, such as Fig. 15B Mapping paired sequence reads from one individual to a reference is the most commonly used sequence-based method for identifying differences in contiguous nucleic acid or genomic structure such as inversions, deletions, and duplications (Tuzun et al., 2005). Fig.15A and Fig. 15BIt is shown how read pairs generated by proximity ligation of DNA from reassembled chromatin of GM12878 were mapped to the human reference genome GRCh38, revealing two such structural differences.

[0343] exist Figure 16A-16C In Figure 2, illustrations of exemplary sequencing differences observed between mapped read pair data and reference scaffolds (right) and images depicting these events (left). Fig.16A In , displaced segment differences are observed, where the scaffold position is mapped to a large number of positions on a single axis (as thin horizontal or vertical lines). The vertical line above the figure indicates the position of the displaced segment, and the arrow then indicates the correct placement of that vertical band in the scaffold. Optionally, the model is updated by repositioning the fragment corresponding to the displaced segment to its correct position in the scaffold. Fig. 16B In , a collapsed fragment situation is observed where fragments A and A' are very similar and map together, but fragments B and B' are very dissimilar (top right), resulting in a scaffold that incorrectly orders the fragments as AB-B' (bottom right). This deviation is identified by an off-diagonal region of unexpected low read density in the image generated by the mapped read pairs (left, region above B'), alternatively or additionally by higher than expected read density near the axis of fragment A (indicating two copies associated with B / B'). If fragments B and B' were arranged in the order suggested by the scaffold (adjacent), then read density corresponding to this adjacency would be expected near the diagonal axis, as observed between the AB fragments. In addition, higher than expected density is observed in the region corresponding to A-B', further suggesting that B and B' are independently adjacent to A, but not to each other. Optionally, the model is corrected by moving B' to a different chromosome, duplicating A on that chromosome, and updating the copy number. In Fig. 16C In the image, folded duplications and misconnections were observed, in which two fragments A and Y were each adjacent to a highly similar sequence B / X, but A and Y were present on different chromosomes. The generated scaffold incorrectly arranged the fragments as A-(B / X)-Y, where B / X was folded and AY was misconnected. This deviation can be identified from the mapped read pair data in the image (left figure), where areas of abnormally low read density are observed on either side of the diagonal axis, but other low-density lines extend outward from the feature at a 45-degree angle from the diagonal axis. Alternatively or in combination, the difference is also identified by an area of ​​higher than expected read density near the axis, corresponding to two copies of B / X relative to A or Y. Optionally, the model is corrected by disconnecting B / X and Y, then copying B / X and attaching it to Y.

[0344] exist Fig.17AIn , an exemplary workflow for improving the quality of mapped read pair data (model optimization) is observed, comprising the steps of obtaining raw connection density data, generating contact likelihood scores, performing side graph editing, generating distance fields, and updating contact likelihoods relative to the current side graph. In some cases, the process results in an interactively updated graph-based genome model. In some cases, the process is iterated to improve the quality of mapped read pair data for feature identification. In Fig. 17B In , an image of raw junction density read pair data mapped onto scaffolds before model optimization on potato chromosomes is observed. Fig. 17C In Figure 1, the same image of read pair data mapped onto a scaffold after model optimization for a potato chromosome is observed. In some cases, the resulting image has fewer off-axis regions of local high density and low density, indicating a better fit of the scaffold model to the read pair data.

[0345] exist Figures 18A-18D In the embodiment of the computer system or network for implementing the method described herein, it is observed. For example, Fig.18A An exemplary computer system that is programmed or otherwise configured to implement the methods provided herein is shown. Fig.18B An example of a computer system that can be used in conjunction with an exemplary embodiment of the present invention is observed in FIG. Fig.18C In FIG. 7 , a block diagram is shown illustrating a first exemplary architecture of a computer system 700 that may be used in conjunction with exemplary embodiments of the present invention. Fig.18D In FIG. 2 , a schematic diagram is observed showing a network 2100 configured to incorporate multiple computer systems, multiple cell phones and personal data assistants, and network attached storage (NAS) that can be used in conjunction with exemplary embodiments of the present invention. Fig.18E , a block diagram of a multiprocessor computer system 900 using a shared virtual address memory space that can be used in conjunction with exemplary embodiments of the present invention is observed. In some cases, computer systems and networks perform the methods described herein without user supervision.

[0346] definition

[0347] As used herein and in the appended claims, the singular forms "a", "an", and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to "a contig" includes a plurality of such contigs, and reference to "detecting the physical layout of a chromosome" includes reference to one or more methods for detecting the physical layout of a chromosome and equivalents thereof known to those skilled in the art, and so forth.

[0348] Likewise, the use of "and" means "and / or" unless otherwise specified. Similarly, "comprising," "including," and "containing" are interchangeable and are not intended to be limiting.

[0349] It should be further understood that when the term "comprising" is used in the description of various embodiments, those skilled in the art will understand that in some specific cases, the embodiments may be alternatively described using the language "consisting essentially of" or "consisting of."

[0350] The term "sequencing read" as used herein refers to a DNA fragment in which the sequence has been determined.

[0351] As used herein, the term "contig" refers to a contiguous region of a DNA sequence. A "contig" can be determined by any number of methods known in the art, such as by comparing sequencing reads of overlapping sequences, and / or by comparing sequencing reads to a database of known sequences to identify which sequencing reads have a higher probability of being contiguous.

[0352] As used herein, the term "subject" may refer to any eukaryotic or prokaryotic organism.

[0353] As used herein, the term "naked DNA" can refer to DNA that is substantially free of complexing proteins. For example, it can refer to DNA that is complexed with less than about 50%, about 40%, about 30%, about 20%, about 10%, about 5%, or about 1% of the endogenous proteins found in the nucleus.

[0354] As used herein, the term "reconstructed chromatin" may refer to chromatin formed by complexing nucleic acid binding moieties with nucleic acids such as naked DNA. In some cases, these moieties are nucleic acid proteins, such as nuclear proteins or histones, but other moieties, such as nanoparticles, may also be considered.

[0355] As used herein, the term "read pair" or "read-pair" may refer to two or more elements that are connected to provide sequence information. In some cases, the number of read pairs may refer to the number of mappable read pairs. In other cases, the number of read pairs may refer to the total number generated.

[0356] As used herein, "tissue sample" refers to a biological sample from an individual or an environment that may contain nucleic acids. For example, a tumor is considered a tissue and a sample collected from a tumor constitutes a tissue sample, but in some cases, the term refers to a sample collected from a heterogeneous environment, such as a stomach or intestinal section, or an environmental sample containing nucleic acids from multiple sources that are spatially distributed relative to each other.

[0357] As used herein, "about" for a number refers to that number + / - 10% of that number. As used herein, "about" for a range refers to a range where the lower limit is 10% less than the indicated lower limit of the range and the upper limit is 10% greater than the indicated upper limit of the range.

[0358] As used herein, "probe" refers to a molecule that transmits information by binding to a target. Exemplary probes include oligonucleotide molecules and antibodies. Oligonucleotide molecules can act as probes by annealing with a target, and transmit information by changing fluorescent characteristics or by annealing with a target and promoting the synthesis of a product (such as an amplicon indicating the presence of a target). That is, the term probe used herein refers to antibody probes and other small molecule probes, and oligonucleic acid molecules, which either directly hybridize with a target to cause, for example, a change in fluorescent state to generate a signal, or act by promoting the synthesis of an amplicon indicating the presence of a target.

[0359] As used herein, the phrase "at least one" when followed by a series such as "A, B, C, D", refers to a single member of the series (A or B or C or D), two members of the series, three members of the series, up to and including all members of the series (A, B, C, and D), and in some cases, other unlisted members. "At least one" in a series does not necessarily mean that every member in the series has a representative.

[0360] As used herein, a DNA protein complex is disrupted or broken when the protein and nucleic acid no longer assemble to form a complex. In some cases, the complex is completely denatured or decomposed so that protein DNA binding is not retained. Alternatively, in some cases, the DNA protein complex is substantially disrupted when the first nucleic acid segment and the second nucleic acid segment are no longer held together independently of any phosphodiester bonds.

[0361] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present disclosure belongs. Exemplary methods and materials are now described, but methods and reagents similar or equivalent to any methods and reagents described herein can also be used in the practice of the disclosed methods and compositions.

[0362] The following examples are intended to illustrate rather than limit the present disclosure. Although they are typical of those that might be used, other procedures known to those skilled in the art may alternatively be used.

[0363] Example

[0364] Example 1.

[0365] A sample comprising three chromosomes is suspected of having at least some genomic material that has undergone at least one genomic rearrangement relative to a reference scaffold. The sample comprises a first chromosome having segments a and b, a second chromosome comprising segments c, d and e, and a third chromosome comprising segments f and g.

[0366] Read pair information for the sample is obtained, and the read pairs are mapped relative to the reference scaffold.

[0367] A local density variation was observed that represented a large overrepresentation of read pairs mapping to segments A and D. It was concluded that a rearrangement occurred that physically connected A and D to each other.

[0368] The local density variation was further analyzed in detail. It was observed that at the peak density of this local density variation, the read pair bin occupancy as a density measure matched the read pair density immediately outside the axis. It was concluded that segments A and D were adjacent in at least one rearrangement event.

[0369] It was observed that the local density variation has symmetry. It was observed that at the resolution level of the mapping, the local density variation is essentially bilaterally symmetrical along the line that bisects the right angle closest to the scaffold axis. It was observed that the translocation includes a segment of a and d that is at least as long as the resolution level of the assay. It was concluded that the event is a simple translocation, resulting in a being adjacent to d.

[0370] Example 2.

[0371] A sample comprising three chromosomes is suspected of having at least some genomic material that has undergone at least one genomic rearrangement relative to a reference scaffold. The sample comprises a first chromosome having segments a and b, a second chromosome comprising segments c, d and e, and a third chromosome comprising segments f and g.

[0372] Read pair information for the sample is obtained, and the read pairs are mapped relative to the reference scaffold.

[0373] A local density variation was observed that represented a large overrepresentation of read pairs mapping to segments A and D. It was concluded that a rearrangement occurred that physically connected A and D to each other.

[0374] Further detailed inspection of the map revealed that segments A and D were not involved in any other substantial off-axis local density variations, concluding that segments A and D were adjacent in a single rearrangement event.

[0375] Example 3.

[0376] A sample comprising three chromosomes is suspected of having at least some genomic material that has undergone at least one genomic rearrangement relative to a reference scaffold. The sample comprises a first chromosome having segments a and b, a second chromosome comprising segments c, d and e, and a third chromosome comprising segments f and g.

[0377] Read pair information for the sample is obtained, and the read pairs are mapped relative to the reference scaffold.

[0378] A local density variation was observed that represented a large overrepresentation of read pairs mapping to segments A and D. It was concluded that a rearrangement occurred that physically connected A and D to each other.

[0379] The map was examined in further detail. It was observed that d was involved in other substantial off-axis local density variation. It was observed that segment d was involved in local density variation that was complementary to the read pairs mapped to g. It was concluded that segments d and g were involved in a rearrangement event that physically connected them.

[0380] Further detailed analysis of the local density variation showed that at the peak density of the local density variation from d to g, the read pair bin occupancy as a density measure matched the read pair density immediately outside the axis. It was concluded that segments d and g were adjacent in at least one rearrangement event.

[0381] The map was examined in further detail. It was observed that a was involved in other substantial off-axis local density variation. It was observed that segment a was involved in local density variation that was complemented by read pairs that mapped to g. It was concluded that segments d and g were involved in a rearrangement event that physically connected them.

[0382] The local density variation was further analyzed in detail. It was observed that at the peak density of the local density variation from a to g, the read pair bin occupancy as a density measure was significantly lower than the read pair density immediately outside the axis. It was concluded that segments a and g were non-adjacent in at least one rearrangement event.

[0383] A more detailed examination of the local density variations of AD and DG observed that each lacked bilateral symmetry along a line drawn from the right angles closest to the axis. It was concluded that a translocation of segment D occurred that was within the resolution level of the map.

[0384] Example 4.

[0385] A sample comprising three chromosomes is suspected of having at least some genomic material that has undergone at least one genomic rearrangement relative to a reference scaffold. The sample comprises a first chromosome having segments a and b, a second chromosome comprising segments c, d and e, and a third chromosome comprising segments f and g.

[0386] Read pair information for the sample is obtained, and the read pairs are mapped relative to the reference scaffold.

[0387] A local density variation was observed that represented a large overrepresentation of read pairs mapping to segments A and D. It was concluded that a rearrangement occurred that physically connected A and D to each other.

[0388] The local density variation was further analyzed in detail. It was observed that at the peak density of this a to d local density variation, the read pair bin occupancy, as a density measure, was approximately half of the read pair density immediately outside the axis. It was concluded that segments a and d were adjacent in at least one rearrangement event.

[0389] The map was examined in further detail. It was observed that d was involved in other substantial off-axis local density variation. It was observed that segment d was involved in local density variation that was complementary to the read pairs mapped to g. It was concluded that segments d and g were involved in a rearrangement event that physically connected them.

[0390] The local density variation was further analyzed in detail. It was observed that at the peak density of this d to g local density variation, the read pair bin occupancy, a density measure, was approximately half of the read pair density immediately outside the axis. It was concluded that segments d and g were adjacent in at least one rearrangement event.

[0391] The map was examined in further detail. Local density variations were observed that were not involved in any of the complements with read pairs mapping to g. It was concluded that segments a and g were not involved in a rearrangement event that physically connected them.

[0392] A more detailed examination of the local density variations of AD and DG observed that each displayed bilateral symmetry along a line drawn at right angles from the nearest axis. It was concluded that a translocation of segment D occurred at a level greater than the resolution of the map.

[0393] The conclusion is that the translocation event linking a to d occurred on one chromosome, and a separate translocation linking d to g occurred on a second chromosome. The conclusion is that the sample is heterozygous for each translocation event.

[0394] Example 5. Conversion of read-pair intervals to kernels

[0395] Read pair data from human chromosome 7 (15 Mb) were obtained, the read pairs were organized into 200 bp bins, and LLR values ​​were calculated for each bin. High LLR values ​​corresponding to known heterozygous inversions were obtained ( Figure 5A In the same analyzed region, a 10 Kb heterozygous deletion kernel and a 26 Kb heterozygous duplication (L) kernel were identified (respectively Figure 5B and Figure 5C ).

[0396] Example 6. Identification of displacement segments

[0397] Read pair information for the sample is obtained and the read pairs are mapped relative to the reference scaffold. Local density variations representing potential misplaced segments of read pairs mapped to segments of the scaffold are observed as vertical or horizontal bands with abnormally high read density ( Fig.16A ). The corresponding horizontal or vertical band with abnormally low read density "hole" is identified, and the expected read pair density of the band is compared with the read pair density of the misplaced segment. The expected read pair density of the hole matches the observed density of the band, and it is concluded that the misplaced segment corresponds to the hole. The scaffold model is adjusted by exchanging the misplaced segment with the hole to generate an improved model.

[0398] Example 7. Identification of folded segments in diploid genomes

[0399] Read pair information for the sample is obtained and the read pairs are mapped relative to the reference scaffold. For a portion of scaffold AB-B', a first region with higher than expected density is observed near the central axis of segment A relative to at least one other region near the central axis. A second region with abnormally low read density is also observed, in some cases appearing as a square or rectangular shape of low density dividing the two segments ( Fig.16A ), one corner of the second region touches the central axis between B and B'. The "excess" density in the first region is approximately proportional to the combined density corresponding to the lack of density observed in the second region. It is concluded that the first region corresponds to a diploid sequence of A that is folded due to high similarity, and the lack of density on or near the axis between B and B' indicates that an incorrect joining has occurred. Optionally, the scaffold is adjusted by duplicating A (increasing the copy number) and disrupting B-B' to produce two independent chromosomes containing AB or A-B'.

[0400] Example 8. Identification of folded repeat sequences and rejoining in diploid genomes

[0401] Read pair information for the sample is obtained and the read pairs are mapped relative to the reference scaffold. For a portion of scaffold AB / XY, a first region of higher than expected density is observed near the central axis of segment B / X relative to at least one other region (e.g., segment A or Y) near the central axis. In addition, a second region of abnormally low read density is also observed, in some cases manifested as a square or rectangular shape of low density ( Fig. 16B ), one corner of the second region does not fully contact the central axis between A and Y. It is concluded that the second region corresponding to B / X contains a folded segment and that A and Y are incorrectly joined by the common segment B / X. Optionally, the scaffold is adjusted by duplicating B / X and disrupting BY to generate two independent chromosomes containing AB or XY.

[0402] Example 9. Identification of chromosome breaks

[0403] Read pair information for the sample is obtained and the read pairs are mapped relative to the reference scaffold. For a portion of the scaffold, lower than expected read density on and off the centroid is observed for the region corresponding to the junction between two segments. It is concluded that there is a chromosome break and the scaffold is updated accordingly.

[0404] Example 10. Identification of haploid folded segments

[0405] Obtain the read pair information of haploid genome sample, and read the pair is mapped relative to reference support.For a part of support, for the region corresponding to the connection between two segments, it is observed that the read density higher than expected on the central axis (for example, higher than the average read density of other regions near the axis on the support).Other important off-axis features are not found.The conclusion is that the high-density region represents the repeated segment folded during the support assembly process.The repeated segment is copied and placed near the original segment in the support.Optionally, iteratively adjust the model until the read density near the axis at the repeated segment is close to the average read density of the position along the support, indicating that the correct number of repeated segments exists in the support model.

[0406] Example 11. Genome modeling

[0407] Read pair information for a tumor sample is obtained, and the read pairs are mapped relative to a human genome reference scaffold. A large number of deviations are observed between the scaffold and the read pair data, manifested as variations between expected and observed densities in multiple regions, which complicates the analysis. A score is given to each deviation based on the magnitude of the deviation. The scaffold is remodeled as a collection of weighted genomes, each containing weighted chromosomes, and the read pair data is remapped. This results in a significant reduction in the number of deviations, and thus a reduction in the score. As a result, data analysis proceeds normally, and information about the heterogeneity of the tumor cell population is obtained. Optionally, the model is iteratively adjusted to further reduce the score and obtain a better fit to the scaffold's read pair data, such as Fig.17A exemplified.

[0408] Example 12. Graphical representation of a scaffold

[0409] Obtain the read pair information of the sample, and the read pair is mapped relative to the reference support. The segments of the support are mathematically represented as nodes, and the regions of the mapped read density are represented as the edges connecting the nodes. Optionally, based on the observed region and the position of the read density, each edge is weighted (e.g., blocking the edge) according to the correct possibility of the connection between the segments. A computational algorithm is used to iteratively evaluate the path through the node along the edge until the shortest path is identified. Optionally, a machine learning algorithm is used to find the shortest path through the graph. The conclusion is that the shortest path represents the best fit support model for reading the data. Representing the assembly support as a graph in this way results in an overall reduction in the computational time and energy required to generate the best fit support model.

[0410] Example 13. Diploid inversion

[0411] A sample comprising a diploid genome is suspected of having at least some genomic material that has undergone at least one genomic rearrangement relative to a reference scaffold. The sample comprises a first chromosome having segments a, b, and c, and a second chromosome comprising segments d, e, and f.

[0412] Read pair information for the sample is obtained, and the read pairs are mapped relative to the reference scaffold.

[0413] Local density variation was observed, representing a large underrepresentation of read pairs mapped to segments ab and bc. It was concluded that a rearrangement (inversion) occurred that connected a and the right end of b, and the left end of b and c.

[0414] The local density variation was further analyzed in detail. It was observed that at the peak density of this local density variation, the read pair bin occupancy as a density measure was only half of the read pair density immediately off-axis. In addition, the displacement density showed a "bow-tie" pattern located off-axis, at the midpoint between segments b. It was concluded that the inversion occurred on only one chromosome.

[0415] The local density variation was observed to impart symmetry. It was observed that at the resolution level of the mapping, the local density variation was essentially bilaterally symmetric along a line bisecting the right angle closest to the scaffold axis. It was concluded that the event was a simple inversion, reversing the orientation of segment b.

[0416] Example 14. Diagnostic Methods

[0417] Tumor samples were collected from patients, sequenced to obtain read pair data, and the resulting data were mapped to a human reference genome scaffold. The methods and systems herein were used to identify off-axis "bow-tie" density features and these features were identified as translocations between genes ETV6 and NTRK3 on one or both chromosomes to form fusions, such as Figure 7As shown. The difference between the expected density and the observed density of the feature indicates the percentage of chromosomes in the tumor cell genome with mutations. Based on the result and optionally reading additional features present or absent in the data, the patient is diagnosed with cancer, such as breast analog secretory carcinoma, and is subsequently treated with drugs (such as NTRK3 kinase inhibitors) known to target cancers with the mutation. Sequencing of samples removed from tumors after completing the treatment regimen indicates a reduction or elimination of the density of features corresponding to the ETV6-NTRK3 translocation. Clinicians conclude that drug treatment has successfully killed tumor cells with translocations in the genome.

[0418] Example 15. Diagnostic Method

[0419] Tumor samples are collected from patients, sequenced to obtain read pair data, and the resulting data is mapped to a human reference genome scaffold. Using the methods and systems herein, no off-axis "bow-tie" density features corresponding to translocations between genes ETV6 and NTRK3 are observed for one or both chromosomes. Based on this result and optionally additional features present or absent in the read pair data, the clinician concludes that the patient does not need to be treated with a drug such as an NTRK3 kinase inhibitor.

Claims

1. A computer-implemented method, the method comprising: (a) mapping read pair sequence information onto a sequence scaffold, wherein mapping the read pair sequence information onto the sequence scaffold comprises positioning a symbol indicative of a read pair such that a distance of the symbol from an axis representing the sequence scaffold indicates a distance from a mapping position of a first read of the read pair on the sequence scaffold to a mapping position of a second read of the read pair on the sequence scaffold, and such that a position of the symbol relative to the axis representing the sequence scaffold indicates an average of the mapping position of the first read of the read pair and the mapping position of the second read of the read pair; (b) identifying local variation in density of symbols of the plurality of read pairs, wherein identifying local variation in density comprises identifying regions having a local low density of symbols and identifying regions having a local high density of symbols; and (c) assigning the local variation of the density to corresponding structural arrangement features, wherein the corresponding structural arrangement is selected from gene truncation, deletion, fusion, insertion, inversion and duplication, or reconstructing the sequence scaffold to generate a reconstructed sequence scaffold, thereby reducing the local variation of the density, wherein the reconstruction includes reordering at least some contigs of the sequence scaffold, redirecting at least one contig of the sequence scaffold, introducing breaks into at least one contig of the sequence scaffold, or translocating a segment of a first contig into an internal region of a second contig, and repeatedly repeating (a) and (b) using the reconstructed sequence scaffold.

2. The method of claim 1, comprising assigning local variations in the density to corresponding structural arrangement features.

3. The method of claim 1, comprising remodeling the sequence scaffolds to reduce local variation in the density.

4. The method of claim 2, wherein reconstructing the sequence scaffold comprises reordering at least some contigs of the sequence scaffold.

5. The method of claim 2, wherein reconstructing the sequence scaffold comprises reorienting at least one contig of the sequence scaffold.

6. The method of claim 2, wherein reconstructing the sequence scaffold comprises introducing a break into at least one contig of the sequence scaffold.

7. The method of claim 6, further comprising introducing a sequence present at one edge of the break onto a second edge of the break.

8. The method of claim 1, wherein reconstructing the sequence scaffold comprises translocating a segment of a first contig into an internal region of a second contig.

9. The method of claim 1, wherein identifying local variation in density comprises identifying a density at a first position and a density at a second position, wherein the density at the first position and the density at the second position are significantly different.

10. The method of claim 9, wherein the first location and the second location are adjacent. The method of claim 9 , wherein the first position and the second position are equidistant from the sequence support.

12. The method of claim 1, wherein identifying local variation in density comprises obtaining an expected density at a first location and an observed density at the first location.

13. The method of claim 12, wherein the expected density at the first position is a density predicted by a density gradient that decreases monotonically with increasing distance from the axis representing the sequence scaffold.

14. The method of claim 1, wherein the scaffold represents a transgenic cell genome.

15. The method of claim 1, wherein the scaffold represents a gene-edited genome.

16. The method of claim 3, wherein the scaffold has an N50 increased by at least 20% after the remodeling.

17. A computer-implemented method, the method comprising: (i) obtaining a scaffold containing sequence scaffold information; (ii) obtaining paired read information; (iii) arranging the paired read information such that at least some of the read pair information is depicted as symbols to indicate a position of each read in a read pair relative to the scaffold and to indicate a distance from one read of the read pair to the other read mapped on the scaffold; (iv) identifying a local variation in density of the paired reads, wherein identifying a local variation in density comprises identifying a density perturbation having a density peak at a vertex of a right angle; and The local variation of the density is assigned to a corresponding structural arrangement feature, wherein the corresponding structural arrangement is selected from gene truncation, deletion, fusion, insertion, inversion and duplication, or the scaffold is restructured to reduce the local variation.

18. The method of claim 17, comprising assigning local variations in the density to corresponding structural arrangement features.

19. The method of claim 17, comprising reconfiguring the scaffold to reduce the local variation.

20. The method of claim 17, wherein obtaining a scaffold comprising sequence scaffold information comprises sequencing a nucleic acid sample.

21. The method of claim 17, wherein obtaining a scaffold comprising sequence scaffold information comprises receiving digital information representing a nucleic acid sample.

22. The method of claim 17, comprising obtaining a predicted density distribution for the deployed read pair information.

23. The method of claim 22, wherein the identifying comprises identifying a significant difference between the predicted density distribution and the depicted read pair information density.

24. The method of claim 17, wherein the vertex of the right angle points toward an axis representing the stent.

25. The method of claim 17, wherein obtaining paired-end read information comprises cross-linking unextracted nucleic acids.

26. The method of claim 17, wherein obtaining paired-end read information comprises cross-linking nucleic acids bound in chromatin.

27. The method of claim 26, wherein the chromatin is native chromatin.

28. The method of claim 17, wherein obtaining paired-end read information comprises binding the nucleic acid to a nucleic acid binding moiety.

29. The method of claim 17, wherein obtaining paired-end read information comprises generating reconstructed chromatin.

30. The method of claim 17, wherein reconstructing the sequence scaffolds comprises reordering at least some contigs of the sequence scaffolds.

31. The method of claim 19, wherein reconstructing the sequence scaffold comprises reorienting at least one contig of the sequence scaffold.

32. The method of claim 19, wherein reconstructing the sequence scaffold comprises introducing breaks into at least one contig of the sequence scaffold.

33. The method of claim 32, further comprising introducing a sequence at one edge of the break onto a second edge of the break.

34. The method of claim 19, wherein reconstructing the sequence scaffold comprises translocating segments of a first contig into an interior region of a second contig.

35. The method of claim 17, wherein the scaffold represents a transgenic cell genome.

36. The method of claim 17, wherein the scaffold represents a gene-edited genome.

37. The method of claim 17, wherein the scaffold has an N50 increased by at least 20% after the remodeling.

38. A computer-implemented method for identifying structural rearrangements relative to a sequence scaffold in a sample, comprising: (i) Map read pair sequence information onto sequence scaffolds using symbols; (ii) assigning said symbol to the bin represented by the pixel; (iii) identifying a local density variation of the symbol having a right angle side pointing toward an axis corresponding to the sequence scaffold and having bilateral symmetry along a line bisecting the right angle side, wherein the symbol represents a mapping position of a read pair on the sequence scaffold and a distance separating the read pairs; as well as The sample is classified as having a simple translocation relative to the sequence scaffold comprising a segment length at least as long as the longest furthest mapped read of the local density variation from the translocation site.

39. A computer-implemented method for identifying structural rearrangements in a sample, comprising: (i) Map read pair sequence information onto sequence scaffolds using symbols; (ii) identifying local density variations of the symbols, the local density variations having right-angled sides pointing toward an axis corresponding to the sequence scaffold, wherein a plurality of the symbols represent mapping positions of read pairs on the sequence scaffold and separation distances of the read pairs; (iii) identifying sub-regions of local density variation that break bilateral symmetry along a line bisecting the right-angled edge; as well as (iv) classifying the sample as having a translocation relative to the sequence scaffold comprising a segment lacking a sequence to which the symmetric restoring read pair population is to be mapped.

40. A computer-implemented method for identifying structural rearrangements relative to a sequence scaffold in a sample, comprising: (i) Map read pair sequence information onto sequence scaffolds using symbols; (ii) identifying local density variations of the symbols, the local density variations having right-angled sides pointing toward an axis corresponding to the sequence scaffold, wherein a plurality of the symbols represent mapping positions of read pairs on the sequence scaffold and separation distances of the read pairs; (iii) obtaining the expected read pair density distribution curve; (iv) identifying a scaffold segment to which the read pair sequence information comprising the local density variation is mapped; and (v) repositioning the scaffold segment so that the read pair sequence information containing the local density variation is mapped to the region indicated by the expected read pair density distribution curve to obtain the density of the local density variation.

Images