k-mer Profiling for Bacterial Identification

Author Name(s): Snehal V. Bhange, Hitesh Tikariha, S. S. Dongre, H. J. Purohit
Author Email: bhange_snehal.ghrcemtechcse@raisoni.net


In this paper, the bacterial identification is done using k-mer profiling. The idea is to create a kmer histogram using R programming and map the occurrences of kmers occurring in common within the same bacteria genera of different species and their respective strains. 5-mer is used in this study for proving the hypothesis. The kmer profile generated clearly distinguishes the two bacteria and look similar for closely related strains. Thus a deep screening and profile matching will help in rapid identification of two bacteria by their kmer profile.


Bacterial Identification, Genera, k-mer Profiling


Bacterial identification is the prominent task in research field related to biological sciences. Therefore researchers keeps on finding novel ideas in this area for bacterial identification. In the last two decades many researchers have found many ways which have evolved in identification of bacteria. Of all these method, statisticians have developed another area for bacterial identification using statistics and data analysis. Since the data in biological size is extremely huge and complicated, the task become more difficult. To overcome this difficulties, this paper proposes the k-mer profiling for bacterial identification and henceforth find the even and odds in the various species and respective strains of same bacteria genera. Of all the strategies in bacterial identification, K-mer counting helps to show more flexible results when they are counted in the spectra of 5<k<20. k-mer counting is initial stage in algorithm related to bioinformatics. k-mer counting is very simple and effective means to study the nature of repetition of subsequences in genomic sequences. In this paper, the major inclination is given toward the use of R Programming[1]. Since R provides the feasible platform where manipulations are effectively handled also it provides with different packages which provides many different functions. R can import files of varies sizes and types. Also, package Biomartr helps to download genome [2] directly with less time consumption. K-mer profiling can be done in various forms like histograms, scatterplot etc. Histograms of k-mer frequencies can help in many ways and give valuable insights into the underlying distribution and indicate the error rate and genome size sampled in the sequencing experiments. k-mer counting helps in generalized statistics which may convey much information about the abundance of data. k-mer finds its way in many advance algorithms like KMC Tallymer, Jellyfish [3], BFCounter, DSK, KMC, Turtle and KAnalyze etc. k-mer generates the prominent characteristics of each genome and finds its way to the comparison with other related genome. Thus the subsequences of each genera is equal to the other genera except that the fluctuation is seen in the diverse numbers of each genera taken. The k-mer thus helps in beneficial statistics of occurrences of each genome.


Hence, the above study shows the statistical analysis of the complete genome and helps to analyse certain properties of respective bacteria based on the k-mer counting. This study on a large scale will surely assist in bacterial        identification and categorizing as per their phylogeny.

1,953 total views, no views today

Download PDF File

About the author: admin