Journal article Open Access

Application of KL Divergence for Estimation of Each Metabolic Pathway Genes

Shohei Maruyama; Yasuo Matsuyama; Sachiyo Aburatani


Citation Style Language JSON Export

{
  "publisher": "Zenodo", 
  "DOI": "10.5281/zenodo.1099834", 
  "language": "eng", 
  "title": "Application of KL Divergence for Estimation of Each Metabolic Pathway Genes", 
  "issued": {
    "date-parts": [
      [
        2015, 
        2, 
        1
      ]
    ]
  }, 
  "abstract": "<p>Development of a method to estimate gene functions is<br>\nan important task in bioinformatics. One of the approaches for the<br>\nannotation is the identification of the metabolic pathway that genes are<br>\ninvolved in. Since gene expression data reflect various intracellular<br>\nphenomena, those data are considered to be related with genes&rsquo;<br>\nfunctions. However, it has been difficult to estimate the gene function<br>\nwith high accuracy. It is considered that the low accuracy of the<br>\nestimation is caused by the difficulty of accurately measuring a gene<br>\nexpression. Even though they are measured under the same condition,<br>\nthe gene expressions will vary usually. In this study, we proposed a<br>\nfeature extraction method focusing on the variability of gene<br>\nexpressions to estimate the genes&#39; metabolic pathway accurately. First,<br>\nwe estimated the distribution of each gene expression from replicate<br>\ndata. Next, we calculated the similarity between all gene pairs by KL<br>\ndivergence, which is a method for calculating the similarity between<br>\ndistributions. Finally, we utilized the similarity vectors as feature<br>\nvectors and trained the multiclass SVM for identifying the genes&#39;<br>\nmetabolic pathway. To evaluate our developed method, we applied the<br>\nmethod to budding yeast and trained the multiclass SVM for<br>\nidentifying the seven metabolic pathways. As a result, the accuracy<br>\nthat calculated by our developed method was higher than the one that<br>\ncalculated from the raw gene expression data. Thus, our developed<br>\nmethod combined with KL divergence is useful for identifying the<br>\ngenes&#39; metabolic pathway.</p>", 
  "author": [
    {
      "family": "Shohei Maruyama"
    }, 
    {
      "family": "Yasuo Matsuyama"
    }, 
    {
      "family": "Sachiyo Aburatani"
    }
  ], 
  "version": "10000800", 
  "type": "article-journal", 
  "id": "1099834"
}
31
10
views
downloads
All versions This version
Views 3131
Downloads 1010
Data volume 2.5 MB2.5 MB
Unique views 2727
Unique downloads 1010

Share

Cite as