Self-Organizing Map for Characterizing Heterogeneous Nucleotide and Amino Acid Sequence Motifs

Abstract

A self-organizing map (SOM) is an artificial neural network algorithm that can learn from the training data consisting of objects expressed as vectors and perform non-hierarchical clustering to represent input vectors into discretized clusters, with vectors assigned to the same cluster sharing similar numeric or alphanumeric features. SOM has been used widely in transcriptomics to identify co-expressed genes as candidates for co-regulated genes. I envision SOM to have great potential in characterizing heterogeneous sequence motifs, and aim to illustrate this potential by a parallel presentation of SOM with a set of numerical vectors and a set of equal-length sequence motifs. While there are numerous biological applications of SOM involving numerical vectors, few studies have used SOM for heterogeneous sequence motif characterization. This paper is intended to encourage (1) researchers to study SOM in this new domain and (2) computer programmers to develop user-friendly motif-characterization SOM tools for biologists.

Supervised Kohonen Self-Organizing Maps of Acute Asthma from Air Pollution Exposure

“…We used an 8 × 8 grid in the final model because this size had an optimum model with lower quantization error. The SOM algorithm can have a rectangular, hexagonal or linear structure and uses the structure to organize the neurons [16]. The principal steps in developing an SOM included (a) initialization, (b) updating of codebook vectors, (c) updating codebook vectors using Euclidian distance and (d) computing the quantization error [15].…”

“…The first step in developing an SOM algorithm is deciding on the size of the map or grid. Selecting the right size of the map is important, as bigger maps with a high number of nodes are computationally heavy and may not always achieve sufficient data reduction [ 16 ]. In the current study, we used a size slightly bigger than the size recommended by the rule of thumb.…”