Gonçalves Lab

Gonçalves Lab @TUDelft (CS @EEMCS_TUD). Research on Computational Molecular Biomedicine, Machine Learning, Oncology by @JoanaGoncalves et al.

[Gonçalves Lab @GoncalvesLab]

Looking back on a most important event: @y_i_tepeli masterfully defended his PhD, 16 Jun 2025, on effective strategies for therapeutic target prediction & selection bias mitigation in ML https://t.co/xzObmBma61. Congratulations Dr. Tepeli!

Looking for his next move, snatch him.

[Anshul Kundaje @anshulkundaje]

Honestly feel kinda sad to see so many young scientists adopting and idolizing ultra hype culture. I think people don't really understand the medium and long term consequences to their own credibility and that of science as a whole.

[Gonçalves Lab @GoncalvesLab]

@y_i_tepeli

[Gonçalves Lab @GoncalvesLab]

Looking back on a most important event: @y_i_tepeli masterfully defended his PhD, 16 Jun 2025, on effective strategies for therapeutic target prediction & selection bias mitigation in ML https://t.co/xzObmBma61. Congratulations Dr. Tepeli!

Looking for his next move, snatch him.

[Gonçalves Lab @GoncalvesLab]

Thank you to everyone contributing to this great celebration! Promotor Marcel Reinders, paranymphs Kirti Biharie & Sander Goossens, the whole DBL lab ❤️, family+friends of Yasin, committee Aalt-Jan van Dijk, Boudewijn Lelieveldt, Jeroen de Ridder, Patrick Kemmeren, Wouter Kouw.

[Gonçalves Lab @GoncalvesLab]

TL;DR
Part I: anti-cancer therapy target discovery, synthetic lethality prediction (ELISL) & stratification of oncogene-addicted cohorts (Oncostratifier, collab Iorio's lab).

Part II: mitigation of selection bias in ML via diversity-guided self-training (DCAST, Metric-DST).

[Gonçalves Lab @GoncalvesLab]

Paper by our Sander Goossens! Shows we can learn tumor-relevant mutational signatures of DDR deficiency from gene perturbation screens with mutation profiling after DNA damage. Check out SNMF (supervised NMF)! https://t.co/gEC97tIAdD #DDR #DNARepair #MachineLearning
@eemcs_tud

[Gonçalves Lab @GoncalvesLab]

Paper by our @y_i_tepeli: How to train fairer ML models from data affected by selection bias? Diversity! DCAST learns from diverse vs. most confident samples to avoid confirmation bias via semi-supervised learning. #MachineLearning #FairnessML @eemcs_tud https://t.co/aRlLLBMCMO

[Gonçalves Lab @GoncalvesLab]

Our graduated BSc students, Andrei Tociu & Emiel Witting, will present their theses at BNAIC/Benelearn'23: assessing domain adaptation for mitigating sample selection bias. Great supervision by @y_i_tepeli! #bnaic #benelearn #MachineLearning #fairnessML #responsibleAI @eemcs_tud

[Gonçalves Lab @GoncalvesLab]

PhD on computational methods to uncover genetic diversity in complex disease using large-scale cell "villages" and single-cell molecular profiling of human donors (iCELL Convergence Flagship). https://t.co/6QWjU7wr6b #Bioinformatics #PhD #vacancy #job #genetics #MachineLearning

[Gonçalves Lab @GoncalvesLab]

Hiring twice, join us at @tudelft!
See thread for details/links.
- Postdoc/PhD: multimodal ML for molecular maps of human tissue in health & disease (@_hubmap).
- PhD: computational analysis of individual diversity in complex disease using cell "villages" of human donors (iCELL).

[Gonçalves Lab @GoncalvesLab]

Postdoc/PhD: multimodal machine learning for molecular maps of human tissue in health & disease (@_hubmap): (spatial) (single-cell) multiomics integration, signatures, cross-modal prediction. https://t.co/lNxuAFjlhx #Bioinformatics #MachineLearning #PhD #Postdoc #vacancy #job

[Yasin Tepeli @y_i_tepeli]

👋 Come visit the poster if you are interested in synthetic lethality. #ECCB2020

[Gonçalves Lab @GoncalvesLab]

@ECCBinfo @cancerSL #MachineLearning

[Gonçalves Lab @GoncalvesLab]

Today @y_i_tepeli is presenting poster P209-T at #ECCB2022 on our effort to improve synthetic lethality prediction. Sequence homology <-> related function? We integrate it w/ cancer omics to predict SL pairs, showing effect on patient survival. Preprint: https://t.co/2AB8A6wlkV

[Gonçalves Lab @GoncalvesLab]

Latest by @y_i_tepeli in our quest to improve synthetic lethality prediction. Isn’t sequence homology indicative of related function? We integrate it with cancer omics to predict SL gene pairs. Differences in patient survival suggest therapeutic potential. @cancerSL w/ C. Seale

[bioRxiv @biorxivpreprint]

ELISL: Early-Late Integrated Synthetic Lethality Prediction in Cancer https://t.co/kffZXLcGc2 #bioRxiv

[Gonçalves Lab @GoncalvesLab]

Colm Seale's MSc thesis in #OUP_Bioinformatics! On selection bias in synthetic lethality prediction. What's up with bias in SL data? Are published models robust? How to diagnose, build more resilient models? https://t.co/FGFdR9mtfp with Yasin Tepeli. #MachineLearning @cancerSL