Prot-Class: A bioinformatics tool for protein classification based on amino acid signatures

Abstract

Knowledge about characteristics shared across known members of a protein family enables their identification within the complete set of proteins in an organism. Shared features are usually ex- pressed through motifs, which can incorporate specific patterns and even amino acid (AA) bi- ases. Based on a set of classification patterns and biases it can be determined which addi- tional proteins may belong to a specific family and share its functionality. A bioinformatics tool (Prot-Class) was implemented to examine pro- tein sequences and characterize them based upon user-defined AA composition percentages and user defined AA patterns. In addition the tool allows for the identification of repeated AA patterns, biased AA compositions within win- dows of user-defined length, and the character- istics of putative signal peptides and glycosyl- phosphatidylinositol (GPI) lipid anchors. Prot- Class is general purpose and can be applied to analyze protein sequences from any organism. The Prot-Class source code is available through the GNU General Public License v3 and can be accessed via the Google Code Repository: http://code.google.com/p/prot-class. Keywords: