kras

Efficient server proxying and mocking in Node.js. 💪

A single-cell transcriptomic analysis of endometriosis, endometriomas, eutopic endometrial samples and uninvolved ovary tissues highlights cell populations characteristic of these tissue types. Transcriptional and cellular heterogeneity across tissues suggests novel therapeutic targets and biomarkers for this disease.

Official repository for the OUP published paper: "Computational Design of a Predicted High-Affinity Humanised Nanobody against KRAS Switch-I". Includes the PIA-KRASv2-Nb sequence, 3D models, and validation data.

Churn_Modelling Using Deep Learning (Implemented ANN)

Validated computational drug-discovery pipeline for KRAS in pancreatic ductal adenocarcinoma (PDAC). Four validated targets (G12C/G12D/G12V/G12R), 518,662 gated compounds, three virtual screens, MM-GBSA rescoring, and a self-audit of a generative campaign whose molecules could not be synthesised. All negative results released.

ML-guided drug discovery, drug repurposing and antisense therapeutics against KRAS mutants in pancreatic cancer: ultra-large docking of 13.6B molecules, FDA-drug repurposing across 18 alleles, and ADAR-recruiting antisense guide RNAs. Nextflow + Python. Code and protocols only, no data.

A genetic study of the allele- and tissue- specific interactions of the KRAS alleles.

Mock web service APIs by providing Swagger YAML files. 💪

Evaluating whether ESM-2 embeddings and AlphaFold structure predict covalently ligandable cysteines: CPU-only, leakage-aware, with orthogonal MS validation

Somatic mutations in TCGA lung adenocarcinoma — maftools workflow

The default frontend / management portal for the kras server. 💪

Please cite our paper: Bhyravbhatla N*, Alsafwani ZW*, Thapa I, et al. Targeting Oncomucin-driven Immunosuppression Improves the Efficacy of K-rasG12D Inhibition in Pancreatic Cancer. Gastroenterology. Published online May 25, 2026. doi:10.1053/j.gastro.2026.04.041

Pharmacophore-based virtual screening and ML classification pipeline for KRAS inhibitor identification

Specificity-first protein binder design: every design is scored against its target and its decoys, and ranked on the margin. Pre-registered thresholds, a measured false-accept rate, and failures reported as results.

De novo binder design against the KRAS G12D neoantigen on HLA-A*11:01, discriminating against wild-type peptide and empty MHC. Anthropic × Adaptyv protein design competition, Challenge 5.

Manuscript for the KRAS comutation research project.

LLM-driven multi-agent simulation of the KRAS G12D pancreatic tumor microenvironment — emergent immunosuppression, NK/DC agents, ablation studies. bioRxiv 2026.