About

Hi, I'm Abdullah Hanif. I use machine learning to understand genomes and proteins, and I'm doing a Master's in Bioinformatics at Saarland University.

Why I built MutaScope. Protein language models are usually summarised by a single score. I wanted to see where they predict well and where they don't, so the benchmark results here come with their test setup, their uncertainty and their limits.

Let's talk. I'd love to join a team working on similar ideas, where machine learning meets genomics and biomedicine. If you're working on something similar, I'd love to hear from you.

Visits are counted with GoatCounter, without cookies: the page, the site you came from, your browser, screen size and country. Nothing that identifies you is stored, and a sequence you score is never sent.