Lukas Gerasimavičius

Our @jmarshlab/@MRC_HGU investigation of structural protein stability predictors using deep mutational scanning (DMS) data has been published in Protein Science @ProteinSociety doi.org/10.1002/pro.46…

New review out with @L_Geras @teichlab ! 🧬🔬 We explore how protein structure is improving variant effect prediction. Advances in structural modelling, including AlphaFold, are powering methods like AlphaMissense, PrimateAI-3D, and CPT-1, leading to better accuracy and interpretability. Read more: sciencedirect.com/science/articl…

We have an open postdoctoral position on deep mutational scanning of transcription factor genes @kudlalab @mrc_hgu @varianteffects. Apply by 4th October! jobs.ac.uk/job/DCK493/res…

Excited to share our Perspective on the new AlphaMissense variant effect predictor tool from @GoogleDeepMind @ScienceMagazine science.org/doi/full/10.11…

I'm happy to present you the final version of Matteo's work on predicting functionally important sites in proteins using a machine-learning model that leverages structure and sequences and is trained using 2D-multiplexed data on abundance and activity doi.org/10.1038/s41467…

Empirical force-field scoring functions remain the most useful stability tools for identifying disease variants. Could machine-learned ensembles of top tools improve agreement with experimental variant effect landscapes? Excited about what future stability predictors might bring.

Finally, repeating the correlation benchmark just on the abundance-based DMS data, also including a number of VEPs, we again find that the 'Foldetta' consensus dominates the ranking, even against dedicated variant effect predictors that were designed specifically for the task.

Our @jmarshlab/@MRC_HGU investigation of structural protein stability predictors using deep mutational scanning (DMS) data has been published in Protein Science @ProteinSociety doi.org/10.1002/pro.46…

Update to our 2020 study benchmarking variant effect predictors using deep mutational scanning data, featuring more VEPs and more DMS datasets. @BenLivesey6 biorxiv.org/content/10.110…

Excited to share this new preprint where we finally are able to address the longstanding hypothesis that cotranslational protein complex assembly can decrease the likelihood of dominant-negative mutations @BdonyiMihly biorxiv.org/content/10.110…

New #biorxiv preprint! We identify structural elements at a human disease locus that promote domain-wide interactions and facilitate extreme long range gene regulation. A fantastic collaboration with Liang-Fu Chen, @minheepark8, Tomek Swigut, @Aboettiger and Joanna Wysocka!

... which is another shameless self-promotion of our paper, recently published in @mSystemsJ on the computational tools to dissect pombe metabolism: journals.asm.org/doi/10.1128/ms… 2/2