Bayes.Feature function

Projecting nucleotide sequences into numeric feature vectors using Bayes kernel encoding approach.

This sequence encoding technique was introduced by Zhang et al. (2006) for prediction of splice sites. In this encoding technique, positional frequencies of nucleotides are computed for both positive and negative datasets, which are then used for encoding of any nucleotide sequence of same length. Each sequence of length L can be encoded into a numeric feature vector of length 2L. Both positive and negative classes of sequences are required for sequence encoding.

Bayes.Feature(positive_class, negative_class, test_seq)

Arguments

• positive_class: Nucleotide sequence dataset of positive class, must be an object of class DNAStringSet.
• negative_class: Nucleotide sequence dataset of negative class, must be an object of class DNAStringSet.
• test_seq: Nucleotide sequences to be encoded into numeric feature vectors, must be an object of class DNAStringSet.

Details

The class DNAStringSet can be obtained by using the function readDNAStringSet avialble in Biostrings package of Bioconductor. Here, the sequences must be supplied in FASTA format.

Returns

A numeric matrix of order $m*2n$, where $m$ is the number of sequences in test_seq and $n$ is the sequence length.

References

Zhang, Y., Chu, C., Chen, Y., Zha, H. and Ji, X. (2006). Splice site prediction using support vector machines with a Bayes kernel. Expert Systems with Applications, 30: 73-81.

Author(s)

Prabina Kumar Meher, Indian Agricultural Statistics Research Institute, New Delhi-110012, INDIA

See Also

POS.Feature, MN.Fdtf.Feature

Examples

data(droso)
positive <- droso$positive
negative <- droso$negative
test <- droso$test
pos <- positive[1:200]
neg <- negative[1:200]
tst <- test
enc <- Bayes.Feature(positive_class=pos, negative_class=neg, test_seq=tst)
enc