sequence-to-function modeling

Target Binding and Sequence Prediction With LSTMs

The paper trains LSTM networks to predict protein properties from primary sequences, achieving high accuracy in classifying homeodomain presence and Artemisinin binding.

BASILIScan: a tool for high-throughput analysis of intrinsic disorder patterns in homologous proteins

BASILIScan identifies structurally stable homologues and disordered regions in proteins to improve recombinant expression success.

Improving on hash-based probabilistic sequence classification using multiple spaced seeds and multi-index Bloom filters

The paper introduces a multi-index Bloom filter for alignment-free sequence classification, showing improved sensitivity and efficiency over existing methods in read binning and taxonomic assignment.

NetTCR: sequence-based prediction of TCR binding to peptide-MHC complexes using convolutional neural networks

A sequence-only convolutional network trained on T cell receptor (TCR)-peptide pairs for HLA-A*02:01 can single out TCRs that bind a target peptide among many non-binding ones.

Fast and sensitive protein sequence homology searches using hierarchical cluster BLAST

The paper introduces a method to speed up protein sequence homology searches with minimal loss of sensitivity and specificity by using hierarchical clustering of reference databases.

Multi-scale periodicities in the functional brain networks of patients with epilepsy and their effect on seizure detection

This paper identifies periodic patterns in functional brain networks of epilepsy patients and shows their correlation with seizure onset, suggesting network-based metrics improve seizure detection.

Deep generative models of genetic variation capture mutation effects

This paper introduces DeepSequence, a deep generative model that captures mutation effects in biomolecules by modeling higher-order dependencies beyond pairwise interactions.

Nucl2Vec: Local alignment of DNA sequences using Distributed Vector Representation

The paper introduces Nucl2Vec, a method using Skip-gram and k-nearest neighbor for efficient local DNA sequence alignment, achieving comparable accuracy to BWA-MEM with reduced computation cost.

nsearch: An open source C++ library for processing and similarity searching of next-generation sequencing data

nsearch is an open-source C++ library for processing and similarity searching of next-generation sequencing data, offering performance comparable to existing tools.

Clustering of Circular Consensus Sequences: Accurate Error Correction and Assembly of Single Molecule Real-Time Reads from Multiplexed Amplicon Libraries

The paper develops a pipeline, C3S-LAA, to correct errors and assemble SMRT reads from multiplexed amplicon libraries, showing improved accuracy over standard methods.

De novo profile generation based on sequence context specificity with the long short-term memory network

The paper introduces a de novo amino acid sequence profile generator using LSTM, showing improved performance over existing methods for predicting beta-strands.

Hybrid correction of highly noisy Oxford Nanoporelong reads using a variable-order de Bruijn graph

This paper introduces HG-CoLoR, a method to correct noisy Oxford Nanopore long reads using a variable-order de Bruijn graph, achieving high-quality correction with efficient runtime.

Mantis: A Fast, Small, and Exact Large-Scale Sequence Search Index

Mantis provides a fast, small, and exact sequence search index for large RNA-seq datasets, outperforming existing methods in speed and accuracy.

Minimap2: pairwise alignment for nucleotide sequences

One mapper covers accurate short reads, noisy long reads, full-length cDNA and megabase contigs, running 3-4 times faster than short-read mappers at comparable accuracy.

Independent assessment and improvement of wheat genome assemblies using Fosill jumping libraries

The paper assesses and improves wheat genome assemblies using Fosill jumping libraries, finding that whole genome shotgun assemblies are more accurate than BAC-based ones.

Opportunities And Obstacles For Deep Learning In Biology And Medicine

Deep learning rivals or beats baselines across imaging, health records, genomics and drug discovery but is not yet transformative; scarce labels, privacy and opacity remain hurdles.

HMMER Cut-off Threshold Tool (HMMERCTTER): Supervised Classification of Superfamily Protein Sequences with a reliable Cut-off Threshold

HMMERCTTER improves protein superfamily classification by introducing a reliable score cut-off threshold using high-quality training data and iterative sequence inclusion.

ProtaBank: A repository for protein design and engineering data

ProtaBank is a database for storing and sharing protein design and engineering data, offering tools for analysis and benchmarking predictive models.

GPU-accelerated alignment of bisulfite-treated short-read sequences

This paper introduces Arioc, a GPU-accelerated short-read aligner for BS-seq data, showing improved speed and accuracy compared to existing tools.

GOLabeler: Improving Sequence-based Large-scale Protein Function Prediction by Learning to Rank

GOLabeler improves sequence-based protein function prediction by integrating multiple classifiers in a learning-to-rank framework, showing better performance than existing methods.

SeqStruct : A New Amino Acid Similarity Matrix Based on Sequence Correlations and Structural Contacts Yields Sequence-Structure Congruence

The paper introduces a new amino acid similarity matrix that improves sequence matching by incorporating structural information, resulting in better sequence-structure congruence.

aBayesQR: A Bayesian method for reconstruction of viral populations characterized by low diversity

This paper introduces aBayesQR, a Bayesian method for reconstructing viral quasispecies from sequencing data, showing improved performance on low-diversity viral populations.

Amino acid stylization: a new approach for representing protein sequences and scoring protein alignments

This paper introduces a new method for representing protein sequences and scoring alignments by stylizing amino acids, offering customizable parameterization for substitution analysis.

Holmes-ITS2: Consolidated ITS2 resources and search engines for plant DNA-based marker analyses

The paper develops a comprehensive ITS2 sequence database and search tools for plant species identification, providing a web service for accurate and efficient marker analysis.

Beyond thermodynamic constraints: Evolutionary sampling generates realistic protein sequence variation

This study shows that evolutionary simulation generates more realistic protein sequence variation than current design methods, highlighting the need for realistic sampling schemes.

Improving pairwise comparison of protein sequences with domain co-occurrence

The paper improves pairwise protein sequence comparison by integrating domain co-occurrence information, increasing BLAST hit sensitivity and proteome coverage.

Sequence fingerprints distinguish erroneous from correct predictions of Intrinsically Disordered Protein Regions

The paper distinguishes native IDP regions from predicted ones using sequence features, finding higher asparagine occurrence and specific pentapeptide combinations in native IDPs.

The standard genetic code facilitates exploration of the space of functional nucleotide sequences

The paper shows the standard genetic code is optimal for exploring functional protein sequence space at intermediate time scales compared to random codes.

Acceleration Of Nucleotide Semi-Global Alignment With Adaptive Banded Dynamic Programming

The paper introduces an adaptive banded dynamic programming algorithm for faster nucleotide sequence alignment, achieving 5x speedup over BLAST+ while maintaining similar sensitivity.

Predicting DNA Hybridization Kinetics from Sequence

The paper predicts DNA hybridization rate constants from sequence using a weighted neighbor voting model trained on experimental kinetics data.