Ireland, Sam M; (2021) Predicting and Characterising Zinc Metal Binding Sites in Proteins. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text Thesis.pdf - Accepted Version Download (3MB) | Preview

Abstract

Zinc is one of the most important biologically active metals. Ten per cent of the human genome is thought to encode a zinc binding protein and its uses encompass catalysis, structural stability, gene expression and immunity. Knowing whether a protein binds to zinc can offer insights into its function, and knowing precisely where it binds zinc can show the mechanism by which it carries out its intended function, as well as provide suggestions as to how pharmaceutical molecules might disrupt or enhance this function where required for medical interventions. At present, there is no specific resource devoted to identifying and presenting all currently known zinc binding sites. This PhD has resulted in the creation of ZincBind — a database of zinc binding sites (ZincBindDB), predictive models of zinc binding at the family level (ZincBindPredict) and a user-friendly, modern website frontend (ZincBindWeb). Both ZincBindDB and ZincBindPredict are also available as GraphQL APIs. The database of zinc binding sites currently contains 38,141 sites, and is automatically updated every week. The predictive models, trained using the Random Forest Machine Learning algorithm, all achieve an MCC ≥ 0.88, recall ≥0.93 and precision ≥0.91 for the structural models (mean MCC = 0.97), while the sequence models have MCC ≥ 0.64, recall ≥0.80 and pre- cision ≥0.83 (mean MCC = 0.87), outperforming competing, previous predictive models.

Download activity - last month

Export as CSVJSONXML

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as CSVJSONXML

Archive Staff Only

View Item