sequence-to-function modeling

Informative RNA-base embedding for functional RNA structural alignment and clustering by deep representation learning

This paper develops an informative RNA-base embedding using deep learning to improve RNA structural alignment and clustering, achieving better accuracy than existing methods.

ProteinBERT: A universal deep-learning model of protein sequence and function

A ~16M-parameter protein language model, pretrained on ~106M UniRef90 proteins by denoising sequences and Gene Ontology annotations, rivals larger TAPE models using linear-cost global attention.

AbLang: An antibody language model for completing antibody sequences

AbLang is an antibody-specific language model that restores missing residues in antibody sequences more effectively than existing methods.

Fast and Optimal Sequence-to-Graph Alignment Guided by Seeds

This paper introduces a novel A[*] seed heuristic for sequence-to-graph alignment, achieving faster and optimal alignment by skipping most dynamic programming table cells.

MetaLogo: a heterogeneity-aware sequence logo generator and aligner

MetaLogo generates sequence logos for heterogeneous sequence groups, allowing users to visualize conservation patterns and annotate unknown sequences.

Stitchr: stitching coding TCR nucleotide sequences from V/J/CDR3 information

Stitchr generates full-length coding TCR sequences from V/J/CDR3 information, validated through synthesis and transduction in Jurkat cells, enabling efficient TCR engineering and analysis.

DSResSol: A sequence-based solubility predictor created with Dilated Squeeze Excitation Residual Networks

DSResSol is a deep learning model that predicts protein solubility from sequence, outperforming existing methods by 5-13% in accuracy and identifying key amino acids contributing to solubility.

Gated recurrence enables simple and accurate sequence prediction in stochastic, changing, and structured environments

The paper introduces a recurrent neural network architecture that enables accurate sequence prediction in stochastic and structured environments through gating, lateral connections, and recurrent weight training.

Neural networks to learn protein sequence-function relationships from deep mutational scanning data

This paper develops a deep learning framework to predict protein function from sequence using deep mutational scanning data and demonstrates its ability to design new proteins with improved properties.

Multi-Branch-CNN: classification of ion channel interacting peptides using parallel convolutional neural networks

The paper introduces Multi-Branch-CNN, a parallel CNN method for classifying ion channel interacting peptides, showing improved performance on novel sequences compared to traditional ML algorithms and other CNN variants.

Learning the local landscape of protein structures with convolutional neural networks

This study uses 3D CNNs to predict wild-type amino acids and consensus sequences from local structural context, finding high-confidence predictions correlate with correct outcomes.

Massively parallel interrogation of protein fragment secretability using SECRiFY reveals features influencing secretory system transit

The paper develops SECRiFY, a method to assess secretability of protein fragments in yeast, revealing intrinsic disorder and flexibility as key factors.

ProDCoNN-server: a web server for protein sequence prediction and design from a three-dimensional structure

ProDCoNN-server predicts and designs protein sequences from 3D structures using deep learning models, offering fast and accurate results with visualization tools.

Strobemers: an alternative to k-mers for sequence comparison

The paper introduces strobemers as an alternative to k-mers for sequence comparison, showing they are less sensitive to mutation rates and provide better match coverage.

Optimal Design of Stochastic DNA Synthesis Protocols based on Generative Sequence Models

The paper introduces a black-box algorithm to optimize stochastic DNA synthesis protocols for generating samples from generative sequence models, showing significant improvements in hit rates for protein engineering.

SPOT-1D-LM: Reaching Alignment-profile-based Accuracy in Predicting Protein Secondary and Tertiary Structural Properties without Alignment.

The paper introduces SPOT-1D-LM, a model that achieves alignment-profile-based accuracy in predicting protein structural properties without using multiple sequence alignments.

Efficient generative modeling of protein sequences using simple autoregressive models

The paper introduces simple autoregressive models for efficient generative modeling of protein sequences, showing comparable performance to complex models at lower computational cost.

Effective gene expression prediction from sequence by integrating long-range interactions

Enformer, a sequence-based transformer integrating elements up to 100 kb away, lifts held-out gene expression prediction (correlation 0.81 to 0.85 versus Basenji2) and variant effect prediction.

XENet: Using a new graph convolution to accelerate the timeline for protein design on quantum computers

XENet improves protein design on quantum computers by reducing rotamer counts and problem size through enhanced graph convolution.

Acetobase: database update and reanalysis offormyltetrahydrofolate synthetase amplicon sequencing data from the anaerobic digesters

The paper updates the AcetoBase database with new sequences and taxonomy, improving taxonomic assignments of FTHFS sequences through reanalysis of amplicon data.

PARROT: a flexible recurrent neural network framework for analysis of large protein datasets

PARROT is a deep learning framework for analyzing large protein datasets, capable of classification and regression tasks using only raw protein sequences.

Navigating the amino acid sequence space between functional proteins using a deep learning framework

This study uses a deep learning framework to generate and explore new protein sequences with functional properties by navigating amino acid sequence space.

Protein sequence sampling and prediction from structural data

The paper improves a model for predicting protein sequences from structures, achieving higher accuracy and showing sampled sequences fold into similar structures.

Prowler: A novel trimming algorithm for Oxford Nanopore sequence data

Prowler improves Oxford Nanopore sequence data quality by retaining phase and read length information, leading to better alignment accuracy but slightly reduced assembly contiguity.

Toward More General Embeddings for Protein Design: Harnessing Joint Representations of Sequence and Structure

This paper introduces a data-efficient method for protein embeddings by jointly training on sequence and structure, improving mutation effect prediction through enriched representations.

Uncovering non-random binary patterns within sequences of intrinsically disordered proteins

This paper introduces NARDINI, a computational method to identify non-random binary patterns in intrinsically disordered proteins by comparing sequence patterns to null models.

TAPER: Pinpointing errors in multiple sequence alignments despite varying rates of evolution

TAPER identifies and removes small stretches of error in sequence alignments while preserving most of the data, improving alignment accuracy.

Study of Basic Local Alignment Search Tool (BLAST) and Multiple Sequence Alignment (Clustal- X) of Monoclonal mice/human antibodies

The paper compares BLAST and Clustal-X alignments of monoclonal antibody light and heavy chains to assess sequence homology and identify conserved regions.

Language models enable zero-shot prediction of the effects of mutations on protein function

A single pre-trained protein language model, ESM-1v, scores mutations zero-shot about as well as family-specific DeepSequence across 41 deep mutational scans, with no experimental labels or retraining.

CRISPR-VAE: A Method for Explaining CRISPR/Cas12a Predictions, and an Efficiency-aware gRNA Sequence Generator

This paper introduces CRISPR-VAE, a framework that explains gRNA efficiency predictions and generates efficient gRNA sequences for CRISPR-Cas12a.