Colm J. Ryan

This paper develops a model to predict resistance to synthetic lethal therapies by integrating protein-protein interaction networks with molecular profiles from CRISPR screens.

OrthoGather provides a user-friendly platform for comparative proteomic analysis and Gene Ontology enrichment across species, enabling functional interpretation of protein sets without computational expertise.

The paper develops GuideRefine to identify and remove multi-targeting and off-targeting sgRNAs from CRISPR-KO libraries, improving the accuracy of gene hit lists.

The paper identifies 121 lncRNA driver mutations across 19 cancer types using 12,631 genomes, showing mutations disrupt RNA-binding protein interactions to promote oncogenesis.

The paper develops a machine learning classifier to predict cell-line-specific synthetic lethality between paralog pairs and validates its predictions using independent CRISPR screens.

The paper identifies CDK8 as a third kinase involved in DNA replication origin activation by showing its cooperation with CDC7 and its role in phosphorylating MCM4.

This study investigates how PRC2 depletion in AML affects epigenomic and transcriptional changes, revealing altered chromatin accessibility, H3K27me3 levels, and gene expression patterns that impact cell phenotype and treatment response.

The paper evaluates sequence and structural similarity metrics for predicting shared paralog functions using synthetic lethality and protein-protein interactions.

The paper shows that gene loss in cancer is often compensated by paralog protein abundance changes, stabilizing protein-protein interaction networks and creating therapeutic vulnerabilities.

The paper assesses whether gene regulatory networks can predict gene sensitivity in cancer cell lines and finds that mRNA abundance is a better predictor than inferred regulatory activity.

The paper assesses how antibody reliability affects mRNA-protein correlations in TCGA tumour samples, finding lower correlations with less reliable antibodies.

The paper shows that measurement reproducibility explains much of the variation in mRNA-protein correlation across studies and pathways.

The paper predicts synthetic lethal interactions between paralog pairs in cancer cell lines using machine learning and features from CRISPR screens and molecular profiles.

The paper introduces a model using knowledge graphs to predict kinase-substrate networks, achieving higher coverage and accuracy than existing methods through network-based learning.

The paper investigates how paralog genes contribute to variable essentiality in cancer cell lines and finds that paralogs from whole genome duplications show more variable essentiality due to buffering relationships.