goal: Drug Resistance could be more hampered by developing drugs that actually fit in the sub-strait envelope (i.e. if a virus develops a change to the drugged area, it should also stop working)
takeaway: to design inhibitors, it sticking out (“protrusion”) of the substrate envelope causes easy areas of mutation that will confer Drug Resistance, therefore, design drugs that try to stay within substrate envelope to ensure a higher degree of imperviousness to mutation (i.e. if the envelope changes well the virus is going to not do its job either)
Messing with HIV
- mutations outside the active site (in primary backbone structure) actually caused a large increase in resistance (4 points outside of backbone structure)
- use linreg and other ML methods to take the type of mutation change (hydrogen? binding? etc.) and to find features most important to confer resistance
Messing with COVID MPro
- COVID has many mutations on the
- aligned sequences to figure conserved interactions
- paxlovid nemdesirvir binds strongly to E166, which—though conserved—still could be resistant to resistance