Bioinformatics Projects

BlocksReader

Abstract

We present a tool known as BlocksReader, developed in Matlab, to aid in Information retrieval and data visualization in Bioinformatics. BlocksReader works by taking as input the entire BLOCKS database and storing the protein sequences as vectors. Using one of twelve methods, users can then encode protein sequences from various families in the database, perform analyses, and visualize the data, which can be helpful in scientific discovery such as phylogenetics and Multiple Sequence Alignment. Encoding families of proteins presents the challenge of vectors of very high dimensions. BlocksReader can also be used as a model, which uses data compression techniques to reduce the dimensionality of large matrices. Once the dimensionality is reduced, scientists and students can visualize and compare protein families, while preserving important features. This approach helps us to discover similarities and relationships between different proteins and their functions as well as to perform database search.

BlocksReader is written in Matlab; therefore, you will need the Matlab software or the runtime Environment. If you do not have matlab, feel free to download the Runtime that is compatible with your system below.