Dmitry Korkin

Korkin received an M.S. in applied mathematics with high distinction from Moscow State University in Russia in 1999.⁠[1]

In 2003 he received his PhD from the University of New Brunswick in Canada under Lev Goldfarb, with the dissertation “A new model for molecular representation and classification: Formal approach based on the ETS framework.”⁠[2]

During his postdoctoral work at the University of California, San Francisco, he published a Protein Science paper with Fred P. Davis and Andrej Sali that analysed 1,847 SCOP domain families from the PIBASE database of domain interactions and found that 72% of them had protein-binding sites more localized than expected by chance.⁠[3]

In June 2009 Korkin, then an assistant professor in the University of Missouri Informatics Institute, received a five-year, $613,000 CAREER Award from the National Science Foundation to study molecular mimicry in the soybean cyst nematode computationally, working with Melissa Mitchum at Missouri to verify predicted binding sites experimentally.⁠[4]

On April 11, 2014, Nature Communications published online a study led by Lilia Iakoucheva of UC San Diego with Korkin among the coauthors; the researchers isolated hundreds of splice variants of autism genes from the human brain and screened their protein interactions, and adding the new splice forms doubled the number of interactions in the network.⁠[5]

In August 2014 Korkin joined Worcester Polytechnic Institute (WPI); he had previously been a tenured associate professor at the University of Missouri-Columbia and had done postdoctoral research at UCSF and Rockefeller UniversityThe Rockefeller UniversityThe Rockefeller University is a biomedical research university on Manhattan’s Upper East Side in New York City, devoted mainly to scientific research and graduate education; it grew out of the Rockefeller Institute for Medical Research founded by John D. Rockefeller.Open the full entry. He later succeeded Liz Ryder as director of WPI’s Bioinformatics and Computational Biology (BCB) program.⁠[6]

In October 2015 the NSF awarded WPI $768,000 for a three-year project led by Korkin to use algorithms to analyse vast genomic data and find genome elements shared across species, which he calls LIMEs; he said his team’s “cache-oblivious” algorithms were showing about a thousand-fold improvement in early results.⁠[7]

In October 2017 WPI described how Korkin’s team used Microsoft HoloLens headsets to display protein interaction networks in 3D to study how genes and proteins linked to disorders such as cancer and diabetes interact; he called the system a tool for the interactive retrieval of information.⁠[8]

In February 2020 Korkin, then an associate professor of computer science at WPI and director of its BCB program, and a team of graduate students used a viral genome published by the National Center for Biotechnology Information to build 3D structural models of the novel coronavirus’s (2019-nCoV) major proteins and their interactions with human proteins, releasing the data freely on his lab website; he said anyone could download the models.⁠[9]

On April 1, 2020, Korkin and eight coauthors including Suhas Srinivasan published a paper in Viruses (volume 12, issue 4) that integrated public data on related coronaviruses into structural genomics and interactomics roadmaps of SARS-CoV-2 and inferred its functional differences and similarities with the SARS coronavirus.⁠[10]

WPI said the paper’s preliminary conclusions included that changes in the new virus’s proteins relative to SARS often formed clusters on protein surfaces; that the spike region targeted by human antibodies differs greatly from SARS, suggesting an existing SARS vaccine may not work; and that regions targeted by earlier SARS antiviral candidates remained largely intact, so those drugs might be repurposed.⁠[11]

He appeared on episode 90 of the Lex FridmanLex FridmanLex Fridman is a Soviet-born, US-raised computer scientist and podcast host who works as an MIT research scientist on human-AI interaction and autonomous driving. His Lex Fridman Podcast is known for hours-long conversations with guests including heads of state and technology leaders.Open the full entry Podcast, discussing the biology of COVID-19, SARS and viruses, and how computational methods can help understand viral protein structure and function for antiviral drug and vaccine development.⁠[12]

He returned for episode 153, whose topics included proteins and the spike protein, virus mutations, protein evolution, the origin of life, Joshua Lederberg and the Dendral expert system, and AlphaFold 2.⁠[13]

In 2023 he and Maksim Kitsak and colleagues published a paper in Nature Communications showing that shortest paths in large real networks such as protein-protein interaction networks and the Internet follow latent hyperbolic-geometric rules, allowing shortest-path nodes to be identified even when most links are missing, and applied the method to cellular pathway reconstruction.⁠[14]

In February 2023 a WPI–McLean Hospital team led by Korkin completed a three-year study using artificial intelligence to predict suicide risk in women with certain trauma-related disorders, accurately predicting suicidal history from data on 123 female patients at the hospital; the work was initiated and funded by the Julia Kasparian Fund for Neuroscience Research.⁠[15]

As of April 2023 Korkin held the Harold L. Jurist ’61 and Heather E. Jurist Dean’s Professor of Computer Science chair at WPI, had stepped down as BCB director to begin a sabbatical, and was leading an NIH-funded study of alternative splicing after earlier work with McLean Hospital. He said the cost of computational methods is orders of magnitude smaller than that of experimental methods.⁠[16]

On April 9, 2024, Nature Neuroscience published a study led by Hagen Tilgner of Weill Cornell Medicine with Korkin among the coauthors, which used single-cell long-read sequencing to map mRNA isoforms across brain regions, cell types and developmental stages in the mouse and human brain.⁠[17]

WPI says Korkin leads a four-year, $1.3 million NIH-funded project with Gloria Sheynkman of the University of Virginia that combines high-level computing and laboratory experiments to develop computational tools predicting the stability and function of protein isoforms.⁠[18]

He has also received a $100,000 grant from the M&H Foundation to support his lab members’ cross-disciplinary research.⁠[19]

X: Public account: https://x.com/dmkorkin (opens in a new window)⁠[12]

University of Missouri: Official website: https://missouri.edu/ (opens in a new window)⁠[4]

Worcester Polytechnic Institute: Official website: https://www.wpi.edu/ (opens in a new window)⁠[9]