b2bTools: online predictions for protein biophysical features and their conservation

Area of Science:

• Biochemistry
• Bioinformatics
• Computational Biology

Background:

• Structural biology and molecular dynamics have limitations in capturing all protein behaviors.
• Understanding inherent biophysical properties encoded in protein sequences is crucial.

Purpose of the Study:

• To provide integrated, sequence-based predictions of protein biophysical behavior.
• To identify emergent properties not easily captured by traditional methods.

Main Methods:

• Utilizing FASTA or text input for single sequences to predict dynamics, folding, disorder, and aggregation.
• Analyzing multiple sequence alignments (MSAs) to explore biophysical variation in homologous proteins.
• Employing Gaussian mixture models to identify unusual residues.

Main Results:

• Developed an online server and API (bio2byte.be/b2btools/) for sequence-based biophysical predictions.
• Predictions include backbone/side-chain dynamics, conformational propensities, folding, disorder, and aggregation.
• Results are visualized via interactive plots and available as downloadable JSON/CSV files.

Conclusions:

• The developed tools offer novel insights into inherent protein properties encoded within their sequences.
• These predictions complement structural and dynamics approaches, providing a more comprehensive understanding of protein behavior.
• The platform facilitates exploration of biophysical variation and limits in protein function.