Discover, rate, and verify biology preprints.

FLIP2: Expanding Protein Fitness Landscape Benchmarks for Real-World Machine Learning Applications

FLIP2 adds seven protein fitness datasets and 16 splits for enzymes, light-sensitive proteins, and interactions; simple ridge and zero-shot baselines often rival fine-tuned protein language models.

Excess salicylic acid mediates graft incompatibility by repressing auxin-dependent vascular reconnection in tomato

Salicylic acid (SA) overaccumulates in tomato-pepper graft rootstocks alongside suppressed auxin; exogenous SA phenocopies failed xylem reconnection, while SA-deficient nahG tomato self-grafts also fail.

Inferring disruption of directed graphs using LIKA reveals altered protein phosphorylation networks in schizophrenia

LIKA, a likelihood model over many-to-many kinase-substrate networks, beats KSEA in simulations with shared substrates, recovers EGFR and ABL1 in cell lines, and ranks MAPK3 first in schizophrenia cortex.

Functional pocketomes and complete allosteric maps of five protein kinases

Energy-model maps of five human kinases show conserved distance-dependent decay of inhibitory allostery, but distal allosteric pockets and activating sites diverge between kinases.

Visual Cortical Response Variability in Infants at High Familial Likelihood for Autism

Among infant siblings of autistic children, greater trial-to-trial variability in visual P1 latency at 6 and 12 months tracked higher 24-month cognitive and language scores; mean P1 measures did not.

jazzPanda: spatially aware marker gene detection for imaging-based spatial transcriptomics

Binning transcript and cell coordinates into spatial vectors lets jazzPanda call smaller, spatially concordant marker sets than Wilcoxon or t-tests on Xenium/CosMx data, while modelling replicates and negative-control background.

ATLAS: Population-Level Disease Locus Discovery via Differential Attention in Genomic Language Models

Contrasting a pretrained genomic language model's attention between case and control haplotypes ranked HBB first on chromosome 11 and recovered more known β-thalassemia variants than GWAS, without per-variant tests.

Invasion status stratifies the composition of the human pancreatic cancer perineural niche

Xenium and single-nucleus profiling of 14 human pancreatic tumors shows invaded nerve neighborhoods enriched for classical cancer cells, myCAFs and lipid-handling macrophages but depleted of lymphocytes.

Nutrient environments shape amino acid auxotrophy and cross-feeding

Across 40 carbon-nitrogen environments, six matched E. coli and B. subtilis amino acid auxotrophs show context-dependent auxotrophy, while auxotroph type predicts cross-feeding yield better than environment.

CellAwareGNN: Single-Cell Enhanced Knowledge Graph Foundation Model for Drug Indication Prediction

Adding OneK1K immune cell-type expression edges to an updated PrimeKG modestly improved drug-indication link prediction, AUPRC 0.816 to 0.826 overall and 0.847 to 0.864 for 40 autoimmune diseases.

BRD4 represses developmental and neuronal genes through interaction with PRC1.6

Acute BRD4 degradation and BD2 mutations de-repress bivalent developmental and neuronal genes in human stem cells and organoids, consistent with BRD4 recruitment by H3K14ac/H3K23ac and interaction with PRC1.6.

Transcriptomic timeseries links hepatic gene expression to an early and self-limited systemic response to enteric infection

Matched liver and colon RNA-seq timeseries during Citrobacter rodentium infection in mice reveal an intimin-dependent hepatic cytokine burst at 3 days that tracks serum cytokine kinetics.

Cellular context restricts a promiscuous m6A reader IDR to a single functional effector interface for mRNA decay

Yeast m6A reader Pho92 uses one helical hydrophobic interface in its disordered region to engage Caf40; a single L35A substitution halts m6A-transcript decay in meiosis.

DetectGxT: detecting gene-by-treatment interactions on molecular count phenotypes accounting for allelic additivity

R package DetectGxT fits allelic-additivity nonlinear models with optional donor or polygenic random effects, raising simulated power to detect gene-by-treatment interaction QTLs over linear regression.

Brain structural and genetic correlates of motor coordination and learning behaviours: modelling developmental coordination disorder

In 14 BXD mouse strains, those modelling developmental coordination disorder have relatively larger cortex and smaller cerebellar/subcortical volumes; this axis tracks motor coordination but maps to no single QTL.

A Multicenter Confirmatory Randomized-Controlled Study of rhNRGβ1 Protein Replacement Therapy in a Murine Model of NF2-related Schwannomatosis

Across three German sites and 207 analyzed mice, systemic rhNRGβ1 at 10 µg/kg did not reduce schwannoma nerve thickness or shift Schwann cell differentiation markers.

Mechanically Compliant, Precision-Porous Brain Implants Reduce the Foreign Body Reaction and Guide Regeneration

Brain-stiffness-matched pHEMA/GMA hydrogel scaffolds with 40 µm interconnected pores reduced astrocyte encapsulation and iNOS+ macrophage fraction versus stiff solid rods after 4 weeks in rat brain.

SVkhor: a unified framework for structural variant integration across long-read, short-read, and optical genome mapping data

SVkhor normalizes, merges and integrates long-read, short-read and optical-mapping structural variant calls; on HG002 it retained 98.6% of naive concatenation's true positives with 58.2% fewer false positives.

Omega-seq: ultra-low-background RNA sequencing with faithful molecular counting and precise transcript-end capture

USER-excisable uracil TSOs, a uracil-intolerant polymerase and excess primer confine UMI incorporation to reverse transcription, minimizing background and UMI inflation; a separate Omega-shaped dT primer captures stranded transcript ends.

Changes in aperiodic (1/f slope) activity during a picture-word interference task: Effects of congruency and sequence manipulations

Single-cycle wavelet 1/f EEG slopes in 49 young adults are steeper on incongruent picture-word interference trials and track trial sequence, but only three effects survive family-wise correction.