Protein structure and sequence visualization

Scientific Programming & ML for Structural Biology

Modeling sequence–structure–function relationships for discovery and design.

Interactive workflows used in bioinformatics analyses

Bioinformatics at scale

Reproducible, open workflows for high-throughput genomics and proteomics.

Molecular docking and ML-driven screening

Computational drug discovery

Learning-guided screening, docking, and hit triage with transparent models.

Mission

Developing and applying open, robust computational methods that connect macromolecular sequence, structure, and phenotype. We build tools that are easy to adopt, transparent to evaluate, and ready to reproduce.

Note: This site represents an independent research effort and collaboration hub; it is not a formal PI “lab.”

Representative figure or group photo placeholder — Replace with a representative figure or team photo.

Research Themes

Sequence ↔ Structure ML

Learning representations that align evolution with biophysics to explain and predict function.

Variant Interpretation

Integrating multiplex assays with machine learning for clinically meaningful effect prediction.

Computational Screening

Scalable docking & active learning to focus wet-lab effort where it matters most.

Project Spotlights

ProteinRep — compact embeddings that capture function across families. Python · PyTorch · UniRef · Docs Preprint Code
VEP-Assist — combining MAVE data with interpretable models for variant effect. R · Scikit-learn · MAVEdb · Docs Code
ScreenSmart — ML-guided docking + uncertainty for small-molecule triage. Nextflow · RDKit · OpenMM · Preprint Code

Software

proteinrep

Embeddings for proteins.

pip install proteinrep

vep-assist

Variant effect utilities.

pip install vep-assist

nf-screensmart

Reproducible screening pipeline.

nextflow run you/nf-screensmart -profile docker

Publications

A short selection appears below; see full list in CV or Google Scholar.

BibTeX

Loading publications…

Resources

Data & Code — GitHub org: @wayyne
Open Science — preprints, negative results, and benchmarking configs.
gPDB weekly wwPDB coverage
- gPDB_seqMap_2025.11.10.fasta
  Multi-FASTA of validated experimental constructs for every wwPDB chain (legacy PDB + mmCIF); PLM-ready.
- gPDB_unpMap_2025.11.7.txt
  Chain → UniProt accession mapping for all wwPDB chains (SIFTS + full mmCIF parsing).
RIDAO rapid intrinsic disorder analysis online
- ridao.app
  Combines 6 established disorder predictors + ANCHOR2 for high-throughput disorder analysis.
- fuzdrop.ridao.app
  Implements FuzDrop within RIDAO to enable large-scale LLPS propensity analysis.

Contact

For collaborations and software questions:

Email: [email protected]