anti-biofilm

anti-biofilm predicts the half maximal inhibitory concentration (IC50) activity of small molecules against microbial biofilms to identify potential anti-biofilm agents.

Key Features:

• Machine Learning Techniques (MLTs): Employs multiple machine learning algorithms including Support Vector Machine (SVM), with SVM achieving a Pearson's correlation coefficient of 0.75 on training and testing datasets.
• Data-Driven Training: Models are trained on experimentally validated anti-biofilm compounds with known IC50 values sourced from the aBiofilm resource and validated against an independent dataset.
• Chemical Diversity Analysis: Analyzes diverse chemical structures including furanone, urea, phenolic acids, and quinolines.

Scientific Applications:

• Drug Repurposing: Predicts IC50 activity of existing compounds to identify candidates for repurposing as anti-biofilm agents.
• Antimicrobial Resistance Research: Supports identification of molecules that can inhibit biofilms, addressing factors that contribute to antimicrobial resistance.
• Chemical Space Exploration: Enables exploration of structurally diverse chemical classes to discover novel anti-biofilm compounds.

Methodology:

Machine learning algorithms were trained on experimentally validated anti-biofilm compounds with known IC50 values from the aBiofilm resource, tested using training and testing datasets where SVM reached a Pearson's correlation coefficient of 0.75, and validated against an independent dataset while performing chemical-structure analyses of classes such as furanone, urea, phenolic acids, and quinolines.