FlexoPred: New Tool For Protein Engineering

University researchers have come up with an SVM approach to analyzing proteins that offers an improvement in specificity for identifying drug inhibition target sites without additional significant cost in turnaround time or computing power. The algorithm is able to extract properties of a protein relevant to protein engineering after reading in just the protein sequence. Current algorithms work poorly when no structure information is available. A working prototype has been used to analyze (1) an arc repressor, (2) PVUII Endonuclease, and (3) Erythroproietin, showing significant advantages over typical disorder-based predictors.

Publication: PLOS - Computational Biology

Case No: SD2006-123
Keywords: drug development, protein analysis, protein engineering, directed evolution, software
Inquiries To:  invent@ucsd.edu