Julien Gagneur

Scooby-2 fine-tunes AlphaGenome to predict single-cell genomic tracks, improving correlation and variant effect prediction compared to Scooby and AlphaGenome.

PepPr improves peptide sequencing by integrating peptide priors, enhancing reference and variant peptide identification while maintaining discovery flexibility.

This study evaluates the diagnostic value of long-read genome sequencing in DEE patients, finding it particularly effective for resolving complex structural variants not detected by short-read sequencing.

The paper benchmarks variant effect scores using UK Biobank data on ultra-rare variant carriers, showing coding and splicing scores better predict protein abundance and traits than transcriptional regulation models.

Parnet is a CLIP-seq-based foundation model for RNA sequence representation learning that outperforms existing methods in binding profile prediction and generalizes across tasks.

FENNEC predicts siRNA activity across chemically diverse designs using a neural network ensemble, outperforming existing methods and supporting design optimization.

The paper introduces UbSeRP to map co-translational ubiquitination across the translatome, revealing RQC substrates, biophysical features predicting RQC, and aging-related changes in CTU.

This paper introduces NTVE, a non-destructive transcriptomics method using vesicular export to monitor RNA dynamics in living cells over time.

The paper presents a standardized RNA-seq analysis framework for rare disease diagnosis, identifying molecular diagnoses in 7.7% of previously undiagnosed cases.

The paper uses fibroblast RNA-seq to improve molecular diagnosis in dystonia patients by identifying RNA expression and splicing defects missed by genomic sequencing.

This study identifies homozygous loss-of-function MDGA2 variants in individuals with developmental and epileptic encephalopathy, establishing MDGA2 as a novel gene for this condition.

Modanovo is a de novo peptide sequencing model that handles post-translational modifications, outperforming existing methods and complementing database searches in phosphoproteomics.

PROTRIDER detects protein expression outliers in proteomics data using conditional autoencoders, outperforming existing methods and revealing enrichments of pathogenic variants in outliers.

scooby models scRNA-seq and scATAC-seq data from DNA sequence to capture cell-specific gene regulation and variant effects.

Flashzoi improves the Borzoi model by using rotary positional encodings and FlashAttention-2, achieving faster training and inference with reduced memory usage while maintaining accuracy in genomic analysis.

The paper improves aberrant splicing variant prediction across developmental stages and identifies developmentally restricted variants linked to neurodevelopmental disorders.

This paper introduces nucleotide dependency analysis to identify genomic functional elements, showing it outperforms existing methods in detecting regulatory elements and RNA structures.

FuncRVP improves rare variant polygenic risk scores by integrating functional gene embeddings, enhancing performance for traits with higher heritability.

This paper introduces EpiOut, a tool to detect chromatin accessibility outliers in ALS, showing they correlate with gene expression and protein level aberrations.

The paper uses HiFi long-read sequencing to identify genetic variants in undiagnosed rare disease families, finding novel diagnoses in 13% of cases and candidate variants in 4.3%.

This paper introduces species-aware DNA language models that capture regulatory elements and their evolution across 800 species, showing improved prediction and motif discovery.

This study shows that using fragment counts instead of binarized data improves scATAC-seq analysis methods like clustering and batch integration.

The paper develops AbExp, a model predicting aberrant gene expression across tissues, achieving 12% average precision by integrating variant effects and tissue-specific expression variability.

RBPNet predicts CLIP crosslink count distributions from RNA sequences, outperforming existing methods and enabling variant impact scoring.

The paper shows that aging in killifish brains reduces basic amino acid-rich proteins through translation pausing, not transcription or proteasome issues, revealing a vulnerability in nucleic-acid protein biogenesis.

This paper introduces DeepRVAT, a deep learning model for rare variant association genetics, showing improved gene discovery and risk prediction using whole-exome sequencing data from UK Biobank.

The paper shows codon optimality affects mRNA stability across human tissues, with energy metabolism and age modulating this effect.

The paper introduces the Intron Jaccard Index to improve aberrant splicing detection in RNA-seq, reducing false positives while maintaining sensitivity in rare disease diagnostics.

The paper adapts the ABC model for enhancer-gene assignment, improves accuracy by incorporating TF binding and single-cell data, and introduces the STARE pipeline.

This paper introduces Spectralis, a deep learning method for de novo peptide sequencing that improves sensitivity and precision in tandem mass spectrometry.