A systematic evaluation of the language-of-viral-escape model using multiple machine learning frameworks DOI Creative Commons

Brent Allman,

Luiz Ângelo Vieira, Daniel J. Diaz

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 8, 2024

Abstract Predicting the evolutionary patterns of emerging and endemic viruses is key for mitigating their spread in host populations. In particular, it critical to rapidly identify mutations with potential immune escape or increased disease burden (variants concern). Knowing which circulating are such variants concern can inform treatment mitigation strategies as alternative vaccines targeted social distancing. A recent study proposed that be identified using two quantities extracted from protein language models, grammaticality semantic change. These defined analogy concepts natural processing. Grammaticality intended a measure whether variant viral viable, change escape. Here, we systematically test this hypothesis, taking advantage several high-throughput datasets have become available, also testing additional machine learning models calculating metric. We find viability, though more traditional metric ΔΔ G appears effective. By contrast, do not compelling evidence useful tool identifying mutations.

Language: Английский

A systematic evaluation of the language-of-viral-escape model using multiple machine learning frameworks DOI Creative Commons

Brent Allman,

Luiz Ângelo Vieira, Daniel J. Diaz

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 8, 2024

Abstract Predicting the evolutionary patterns of emerging and endemic viruses is key for mitigating their spread in host populations. In particular, it critical to rapidly identify mutations with potential immune escape or increased disease burden (variants concern). Knowing which circulating are such variants concern can inform treatment mitigation strategies as alternative vaccines targeted social distancing. A recent study proposed that be identified using two quantities extracted from protein language models, grammaticality semantic change. These defined analogy concepts natural processing. Grammaticality intended a measure whether variant viral viable, change escape. Here, we systematically test this hypothesis, taking advantage several high-throughput datasets have become available, also testing additional machine learning models calculating metric. We find viability, though more traditional metric ΔΔ G appears effective. By contrast, do not compelling evidence useful tool identifying mutations.

Language: Английский

Citations

2