deep learning

CMVAE: Contrastive Supervision Improves Single-Omic Latent Manifold for Breast Cancer Survival Prognosis

NEPTUNE: a neuronal cell death pathway prediction tool used to decipher neurodegeneration using single cell RNA-seq datasets

NEPTUNE identifies neuronal cell death markers from single-cell RNA-seq data, revealing non-apoptotic mechanisms in photoreceptor degeneration in rd1 mice.

Local fractal dimension as an indicator of functional reorganization in molecular interaction networks

The paper introduces DIFFRACT, a framework that uses local fractal dimension changes in protein interaction networks to detect functional reorganization, showing its utility in identifying novel biological roles in RAVER1-deficient cells.

Top2All: Benchmarking sparse-to-transcriptome reconstruction from high-expression profiles across sex- and sex-steroid-defined conditions

The paper benchmarks sparse-to-transcriptome reconstruction models using high-expression profiles from mouse adrenal cells under sex and sex-steroid conditions, finding that gene-wise metrics and strong baselines are critical for evaluation.

A new tool for quantifying mitochondrial RNA processing intermediates reveals increased levels of immature mRNA species under oxidative stress

Probing bacterial fitness in virtual chemostats

The paper develops computational algorithms to accurately estimate bacterial population growth rates using single-cell models and applies them to study antibiotic response heterogeneity and chemotaxis effects.

Consensus methods for trees of blobs and level-1 phylogenetic networks

This paper introduces a consensus framework for phylogenetic networks, demonstrating its effectiveness on simulated and real datasets.

A single-cell multimodal foundation model enables cross modality analysis between transcriptome and chromatin

MUTUAL enables cross-modality analysis between transcriptome and chromatin by learning a shared RNA-ATAC space from multiome data without refitting.

Gene regulatory logic underlying the differentiation of mixed lineages in human thymic epithelial cells

The paper identifies 17 human thymic epithelial cell subsets using single-cell data and shows they share a common ancestor and regulatory logic, with some conservation in mice.

scRM: A single-cell foundation model trained on scRNA-seq reads

scRM is a single-cell foundation model trained on raw scRNA-seq reads, showing that raw data pretraining is feasible and cost-effective compared to count-based models.

Predicted condensation-related signatures localize to two C-terminal regions of the prion protein across cervid PRNP variants associated with chronic wasting disease

Quantifying subcellular organization from localization to co-occurrence with InSpace

InSpace quantifies subcellular organization by preserving localization and molecular co-occurrence, revealing spatial programs in B and plasma cells across cancers.

ToxINsight: An Integrated Web Platform for Sequence, Structure, and Ligand Binding Analysis of Snake Venom Toxins

ToxINsight provides a web platform for analyzing snake venom toxin sequences, structures, ligand interactions, and geographical distributions to aid in antivenom development.

End-to-end reconstruction for cryo electron ptychographic single-particle analysis

This paper introduces an end-to-end reconstruction algorithm for cryo-ptycho single-particle analysis, achieving better resolution than sequential methods on simulated data and some experimental datasets.

The transformer as a candidate canonical cortical microcircuit

The paper compares transformer architectures to cortical microcircuits, finding that transformers uniquely capture circuit details, functional maps, and wiring efficiency.

Decoding antibody-antigen recognition: the BL8 language and its application in calibrating deep learning predictions

This paper introduces BL8, a structural language to evaluate deep learning predictions of antibody-antigen interactions, revealing false positives and enabling reliable consensus predictions.

De novo design of mini-proteins as specific binders to PACAP-38, a key mediator of migraine

This study designs mini-proteins that bind PACAP-38 with high affinity, showing improved potency and stability compared to earlier versions.

LIMBER: A Hybrid RNA Foundation Model for Predicting and Understanding RNA Structure and Function

LIMBER is a hybrid RNA foundation model that predicts RNA structure and function, showing strong performance in secondary-structure prediction and classification, along with transferability to related tasks.

Cross-modal transfer enables zero-shot prediction of drug responses across cell types

GRAFT predicts drug responses across cell types without direct measurements by transferring genetic perturbation data to chemical perturbations.

GLOM TGITO: multiscale protein modelling and multiview landscape-guided sequence discovery

The paper introduces GLOM-TGITO, a method for protein design that combines multiscale modeling with multiview landscape-guided optimization, achieving high accuracy and novel sequence discovery.

Differentiable Spin-Center Diffusion Model Enables DEER-Guided Generative Pipeline of Weighted Protein Ensembles

This paper introduces a differentiable spin-center diffusion model to generate weighted protein ensembles from DEER spectroscopy data, enabling thermodynamic interpretation of conformational ensembles.

EMBER: Contextualized gene representations enable compositional prediction across cellular perturbations

EMBER, a gene-centric foundation model, improves cross-context perturbation-response prediction and genetic-interaction ranking using contextualized gene embeddings.

Intra-DNA k-mer Conservation Patterns Encode Evolutionary Selection of Variants

The paper introduces kGain, a metric quantifying k-mer frequency shifts due to mutations, and shows it predicts adaptive potential across species and experimental systems.

Single-cell perturbation prediction near experimental reproducibility with GENESIS

GENESIS predicts single-cell perturbation responses from experimental conditions, outperforming existing models and approaching experimental reproducibility.

Recurrent Spiking Neural Networks for Continuous Motor Control: A Biologically Plausible Model of Motor Cortex

This paper develops a biologically plausible recurrent spiking neural network model of motor cortex that learns continuous motor control through synaptic plasticity and feedback mechanisms.

Small molecule structure elucidation from NMR spectra with generative and reasoning models

This paper introduces an AI system for automated small molecule structure elucidation from NMR spectra, combining generative models with reasoning agents to outperform existing tools.

Reconstructing human gene regulatory evolution using deep learning

This study uses deep learning to predict cis-regulatory effects of human-specific mutations across 305 tissues, revealing their role in gene expression evolution.

Single-cell foundation models transfer across deep evolutionary distances

This paper evaluates the transferability of single-cell foundation models across deep evolutionary distances using developmental atlases of Dictyostelium discoideum and Caenorhabditis elegans.

Neanderthal introgressed ancestry reveals human genomic regions enriched with recessive deleterious mutations

The paper uses Neanderthal ancestry patterns to identify genomic regions enriched with recessive deleterious mutations using a machine learning classifier.

ChAI-Net learns a regulatory code for chromatin co-accessibility

The paper introduces ChAI-Net, a deep learning model that predicts chromatin co-accessibility from DNA sequences in mouse embryonic stem cells, revealing sequence motifs that govern coordinated TF binding.