Login / Signup

Four conserved amino acids on human papillomavirus E6 predict clinical high-risk types.

Akari NiiyaYo HamaguchiHiroyuki MishimaShoko MiuraNahoko KomatsuKoh NagataYuri HasegawaKiyonori MiuraKoh-Ichiro Yoshiura
Published in: Journal of medical virology (2023)
Human papillomavirus (HPV) types included in the genus alpha papillomavirus (alpha-HPVs) are subdivided into high- and low-risk HPVs associated with tumorigenicity. According to conventional risk classification, over 30 alpha-HPVs remain unclassified and HPV groups phylogenetically classified using the L1 gene do not exactly correspond to the conventional risk classification groups. Here, we propose a novel cervical lesion progression risk classification strategy. Using four E6 risk distinguishable amino acids (E6-RDAAs), we successfully expanded the conventional classification to encompass alpha-HPVs and resolve discrepancies. We validated our classification system using alpha-HPV-targeted sequence data of 325 cervical swab specimens from participants in Japan. Clinical outcomes significantly correlated with the E6-RDAA classification. Four of five HPV types in the data set that were not conventionally classified (HPV30, 34, 67, and 69) were high-risk according to our classification criteria. This report sheds light on the carcinogenicity of rare genital HPV types using a novel risk classification strategy.
Keyphrases
  • deep learning
  • machine learning
  • high grade
  • amino acid
  • big data
  • cervical cancer screening
  • gene expression
  • electronic health record
  • artificial intelligence
  • drug delivery
  • cancer therapy
  • genome wide analysis