Hidden Markov Model Analysis of Motifs in interleukins and haematopoietic growth factor family

Haematopoietic cytokines are important in the regulation of haematopoiesis and immune responses, and they can also influence lymphocyte development. Hundreds of members of several different cytokines families have been discovered by some different methods. But fast evolution rate and low similarity of cytokines prevent identifying novel members of a cytokine family completely with classical tools such as BLAST. Here hidden Markov model algorithm is performed on haematopoietic cytokines and interleukin 10 (IL10) related family, and two motifs are discovered. Then, three famous protein sequence databases (SwissProt, IPI and Nr) were scanned using each motif respectively, which led to evaluate the relation between motif and every sequence of database including some unknown proteins or some ambiguous function proteins of known families. Furthermore, the biology features were compared to filter novel members of cytokines. As a result, four proteins are predicted to be the cytokine candidates