7 results about "Sequence dependent" patented technology

The invention relates to the technical field of gene sequencing, in particular to an enzyme composition for a DNA next-generation sequencing library and a library construction method. The enzyme composition comprises Vvn and a high-fidelity Bst DNA polymerase, and is used for carrying out fragmentation treatment on DNA. The Vvn has no sequence preference for fragmentation of double-stranded DNA (dsDNA), so that the homogeneity of the library is improved, and support is provided for the accuracy of subsequent sequencing results and the integrity of genome coverage. Under the polymerization action of the high-fidelity Bst DNA polymerase, the enrichment of the dsDNA is finally realized, and the sequence authenticity of the enriched product is guaranteed to the maximum extent. The synergistic effect of the Vvn and the high-fidelity Bst DNA polymerase does not depend on sequence specific recognition, various dsDNAs can be efficiently enriched, the method is suitable for library construction of low-abundance nucleic acid, such as construction of a DNA sequencing library with the initial quantity as low as 10 pg, the success rate of library construction can be remarkably increased, and the method is particularly suitable for low-initial-quantity scenes such as precious samples or low-concentration DNA samples, forensic trace DNA and single cell sequencing.

The invention discloses a computer-aided antibody off-target effect prediction method and system. The method mainly comprises the following steps: constructing an antibody-antigen structure and function annotation database, carrying out holoproteome structure search on an antibody by adopting a 3Di / aa space structure descriptor, and carrying out multi-dimensional comprehensive evaluation on a search result in combination with structural similarity. According to the method, high-flux and low-cost prediction of the antibody off-target effect can be realized, the limitation that the traditional method only depends on sequence alignment or single structure scoring is broken through, the prediction accuracy is improved, the risk caused by off-target is reduced, and the research and development period of antibody drugs is remarkably shortened.

The invention discloses a distributed flow shop scheduling method based on a large model aided multi-objective optimization algorithm, and particularly aims to solve the problems of insufficient sequence correlation setting time, insufficient wait-free constraint processing and low multi-objective optimization efficiency in existing distributed heterogeneous factory scheduling. And proposing a non-dominated sorting genetic algorithm based on large language model assistance. The method comprises the following steps: constructing a dual-objective optimization model, adopting a one-dimensional integer array coding solution structure and separating a factory operation sequence through '-1'; in combination with a greedy algorithm idea, selecting a greedy initialization algorithm based on maximum completion time or a greedy initialization algorithm based on sequence-dependent setting time to initialize a population; designing a large model cue word including problem definition, solution example, evolution instruction and legality verification, and executing parent selection, crossover and mutation operation by a large language model; and updating the population through non-dominated sorting, and finally outputting a Pareto frontier solution.

According to the invention, the EnzyForge AI deep learning screening method for novel enzyme molecule mining is constructed and provided for the first time. Aiming at the technical bottleneck that traditional enzyme screening depends on sequence homology, and low-homology enzymes with similar functions are difficult to effectively discover, the method creatively combines sequence homology search and structural similarity search, and takes a local catalytic key structural domain as a function judgment core; on the basis, local residue comparison is further carried out, and multi-dimensional evaluation is carried out on the activity, the catalytic potential and the stability of the candidate enzymes by combining a protein large language model, so that deep screening and sorting of the candidate enzymes are realized. Finally obtained protein is cloned to a pET-28a carrier, heterologous expression is realized in escherichia coli BL21 (DE3), a closed-loop process from computational prediction to structure screening to activity evaluation to experimental verification is constructed, and systematic and large-scale intelligent mining of lipase is realized.

The invention provides an Ethereum intelligent contract transaction sequence dependence vulnerability data set generation method based on variation testing. According to the method, firstly, a target contract is translated into a three-address code through static analysis, and then an assertion state dependency function pair is extracted and a transaction sequence dependency lock used for constraining a transaction sequence is positioned; secondly, eliminating assertion statements in the transaction sequence dependency lock by adopting a mutation operator, and damaging the transaction sequence dependency lock, so as to generate a mutation contract possibly containing transaction sequence dependency vulnerabilities; and then, generating a transaction sequence containing assertion state dependency function pair calling for the variation contract by using a dynamic symbolic execution technology, and verifying whether the transaction sequence dependency vulnerability is successfully injected or not by executing and comparing contract states of the two sequences in a local private chain. And finally, analyzing three address code features of the variation contract to determine vulnerability subordinate categories, and finally constructing a high-authenticity transaction sequence dependent vulnerability data set of subdivided categories.

The application discloses a teaching data dynamic classification processing method and device and a storage medium, constructs a composite modeling architecture based on combination of a multi-head attention, a long short-term memory network and a hidden Markov model, extracts global time sequence dependent features through the multi-head attention mechanism, and performs feature-level fusion with local state features, so that the limitation of a single model in structured data modeling is solved. Secondly, a Tra-BiLSTM double-encoding feature enhancement module is proposed, on the basis of a pre-training model, the global interaction advantage of a transformer and the local bidirectional modeling capability of a BiLSTM are combined, multi-level semantic feature extraction of Chinese time sequence data is realized, and the joint capturing capability of long-distance dependence and local patterns is significantly improved. Thirdly, a dynamic weight multi-task optimization mechanism is designed, the gradient conflict and the unbalanced optimization target existing in multi-task learning are solved, and the overall training efficiency and the generalization performance of the model are improved.