Detecting remote protein evolutionary and structural relationships via string scoring method

The amount of information being churned out by the field of biology has jumped manifold and now requires the extensive use of computer techniques for the management of this information. In this work, we propose, an effective learning method for detecting remote protein homology. The proposed method uses a transformation that converts protein domains into fixed-dimensional representative feature vectors, where each feature records the sensitivity of a set of substrings to a previously learned protein domain. These features are then used to compute the kernel matrix that will be used in conjunction with support vector machines. The proposed method is tested and evaluated on two different benchmark protein datasets and it's able to deliver remarkable improvements over most of the existing homology detection methods.