Dmitry Ivankov

Having fun with explaining everything I know to everybody else

[Dmitry Ivankov @DmitryIvankov2]

Thanks to the fantastic mega dataset of @KotaroTsuboyama we with @kluwik assess the unbiased performance of current ∆∆G predictors with unprecedented accuracy (https://t.co/ouLLsRLHei). The limitation of the Pearson correlation of ~0.60 seems to be confirmed.

[Dmitry Ivankov @DmitryIvankov2]

Note also that tons of AlphaFold models do not make the ΔΔG predictions more accurate: the correlation remains to be ~50-60%. Looking forward to AI revolutionizing ΔΔG prediction.

[Dmitry Ivankov @DmitryIvankov2]

Want to predict protein stability change due to mutation but no 3D structure in PDB? With @kluwik we showed that AlphaFold models or homologous structures are almost as good as original structures for ΔΔG predictions! I-Tasser homology models perform worse https://t.co/eyg220x5Ku

[Denis Wirtz @deniswirtz]

Our databases for PhD/postdocs/early-career faculty featured in @nature.

“We had a choice to reserve the information or make it available to everyone,” the dude says. “I reminded myself that the university’s mission is knowledge for the world.”

https://t.co/MV7WM53Peb

[Guillaume Filion @thegrandlocus]

I am very excited to present our work on mapping quality in our latest bioRxiv preprint "Mapping short reads, faithfully" https://t.co/VHWCsDODL7 #Bioinformatics #Algorithms

[Dmitry Ivankov @DmitryIvankov2]

Our review has just been published in the special issue devoted to 90 anniversary of Oleg Ptitsyn: https://t.co/8rucAGcHo5

[Dmitry Ivankov @DmitryIvankov2]

Our paper about hypercubes (preliminary view) appeared on the Bioinformatics site: https://t.co/ozn5Mn2Qtu