Sabi­na Podlewska

The invaluable role of machine learning methods at different stages of drug design pipelines.

The invalu­able role of machine learn­ing meth­ods at dif­fer­ent stages of drug design pipelines

Abstract

Com­put­er-aid­ed meth­ods are nowa­days an essen­tial com­po­nent of the drug dis­cov­ery work­flow, pro­vid­ing use­ful tools var­i­ous stages of this expen­sive and time-con­sum­ing process. Their ser­vice starts at the very begin­ning, when com­pu­ta­tion­al tools are applied to iden­ti­fy new com­pounds with poten­tial­ly desir­able bio­log­i­cal pro­files. How­ev­er, in sil­i­co eval­u­a­tion is not only lim­it­ed to the assess­ment of the activ­i­ties of the inves­ti­gat­ed mol­e­cules towards con­sid­ered tar­gets but also involves analy­sis of their physic­o­chem­i­cal and phar­ma­co­ki­net­ic prop­er­ties and poten­tial tox­i­c­i­ties, as well as gen­er­a­tion of new poten­tial drug candidates.

The pre­sen­ta­tion will sum­ma­rize the areas of appli­ca­tion of machine learn­ing meth­ods in projects imple­ment­ed in the Depart­ment of Med­i­c­i­nal Chem­istry Maj Insti­tute of Phar­ma­col­o­gy Pol­ish Acad­e­my of Sci­ences. They involve both method­olog­i­cal ini­tia­tives aimed at the devel­op­ment of nov­el machine-learn­ing-based algo­rithms, as well as projects ori­ent­ed at the search of struc­tural­ly nov­el lig­ands of select­ed G pro­tein-cou­pled recep­tors (most­ly sero­tonin and opi­oid recep­tors). In addi­tion, the on-line plat­forms allow­ing the usage of the con­struct­ed tools by wide com­mu­ni­ty will be pre­sent­ed. They refer most­ly to the pre­dic­tion of com­pound ADMET prop­er­ties and indi­ca­tion of chem­i­cal moi­eties influ­enc­ing to the high­est extent the val­ue of the con­sid­ered parameter.

Biog­ra­phy

I am cur­rent­ly Assis­tant Pro­fes­sor in the Depart­ment of Med­i­c­i­nal Chem­istry Maj Insti­tute of Phar­ma­col­o­gy Pol­ish Acad­e­my of Sci­ences, where I have worked since 2010. I obtained my PhD (dis­ci­pline: chem­istry) in 2016 and earned habil­i­ta­tion in 2023 (dis­ci­pline: phar­ma­ceu­ti­cal sciences).

After com­plet­ing my PhD, I was head of my three-year post-doc grant SONATI­NA, which I run at the Fac­ul­ty of Phar­ma­cy Jagiel­lon­ian Uni­ver­si­ty Med­ical Col­lege and cur­rent­ly I am prin­ci­pal inves­ti­ga­tor of the grant OPUS, which is imple­ment­ed in IP PAS. So far, I have been PI in five research grants fund­ed by the Nation­al Sci­ence Cen­tre, and I have been mem­ber of the research team in 13 anoth­er projects. My research work result­ed in 47 arti­cles in sci­en­tif­ic jour­nals, 1 book chap­ter, and over 70 con­fer­ence pre­sen­ta­tions. I was also appre­ci­at­ed by the Foun­da­tion for Pol­ish Sci­ence and award­ed the START schol­ar­ship in 2019.

I am chemist and math­e­mati­cian by edu­ca­tion and in my work I focus on the appli­ca­tion of machine learn­ing meth­ods in the process of search­ing for new drugs with the par­tic­u­lar atten­tion paid to the lig­ands of the G pro­tein-cou­pled recep­tors. Recent­ly, the core of my work are explain­abil­i­ty approach­es and on-line appli­ca­tions of the devel­oped pre­dic­tive tools.