Yeona Kang
Howard University, DC·Mathematics & Statistics
Microbes evolve to escape the drugs we use against them, and that threat has existed since penicillin. This project builds a mathematical way to anticipate that escape, using as its test case cabotegravir in long-acting injectable form (CAB-LA) — an HIV drug given as a shot that is used both to treat infection and as pre-exposure prophylaxis (PrEP), medicine taken by uninfected people to prevent HIV.
- Fields
Mathematics & Statistics
Microbiology & Immunology
Computer Science
- Topics
drug resistance
pre-exposure prophylaxis
PrEP
- Grants
- 1 known grant, through Aug 2029
What they work on
- Predict how resistance mutations change cabotegravir binding to HIV by combining molecular docking with physics-informed machine learning models that estimate inhibitor binding energies for wild-type and mutant virus.
- Translate molecular binding predictions into drug levels and antiviral activity over time using an established pharmacokinetic/pharmacodynamic model of long-acting injectable cabotegravir (CAB-LA).
- Couple drug-activity predictions to immunological and evolutionary models of early HIV infection to simulate when resistant viral variants emerge and are selected under clinically realistic CAB-LA exposure.
- Run data-driven virtual clinical trials across dosing schedules and patient scenarios, validating predictions against biochemical inhibition assays, SHIV macaque studies, human cohort data, and reported HIV reservoir and rebound kinetics.
More about the research
Because a single injection keeps drug in the body for months, the drug level slowly falls through a window where it may be too weak to block the virus but strong enough to favor resistant variants. Predicting when that happens could guide safer dosing.
Why it matters
Long-acting injectable PrEP is being rolled out widely, and resistance emerging during the slow decline of drug levels could blunt both prevention and later treatment options. A validated simulation framework could help identify dosing schedules and patient situations that carry the highest resistance risk, informing safer PrEP strategies, and the same modeling approach could be reused for other drugs and pathogens.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Gauri Rao
University of Pittsburgh, PA·Pharmacology
Identify which pharmacokinetic and pharmacodynamic (PK/PD) factors should guide phage selection against clinical Pseudomonas aeruginosa isolates, using static time-kill assays and hollow fiber infection models.
Elizabeth Thompson
Duke University, NC·Pharmacology·Early-career
Identify the best-performing population pharmacokinetic model of milrinone, a drug that helps the heart pump, by externally validating existing models against prospectively collected real-world data from an independent group of critically ill infants.
Matthew R. Reynolds
University of Wisconsin-Madison, WI·Microbiology & Immunology
Determine how long virus takes to rebound after stopping antiretroviral therapy in rhesus macaques given defined reservoirs of 5x10^4, 1x10^4, 1x10^3 or 1x10^2 laboratory-generated, latently SIV-infected autologous cells.
The school
Howard University: 8 labs on PI Pages, $102M a year on research, 102 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.