Документ взят из кэша поисковой машины. Адрес оригинального документа : http://mccmb.belozersky.msu.ru/2015/proceedings/abstracts/147.pdf Дата изменения: Mon Jun 15 15:41:33 2015 Дата индексирования: Sat Apr 9 23:35:24 2016 Кодировка:

The construction of gene networks for Mycobacterium Tuberculosis by analyzing next-generation sequencing data
Misko Vladislav Moscow State Forest University misk.vlad@gmail.com Introduction Determination of bacterial resistance to antibiotics is an expensive and complicated process. To efficiently create new drugs one needs to understand interactions between different genes and communications within the bacterial regulatory network. To solve this problem methods of identification of expressed genes groups are important. Methods and Results We developed an algorithm for assessment of modulons (groups of genes co-expressed in specific conditions) in Mycobacterium Tuberculosis (strain H37Rv). For this purpose we selected a set of expression data obtained with Next-Generation Sequencing from GEO (Gene Expression Omnibus)[1]. To identified clusters for co-expressed genes, we used self-organizing neural networks[2] and analyzed correlation between different groups of experiments. The results of each experiment consisted in the form a vector of expression levels of each gene.

Figure 1. Self organizing maps of gene expression data for different numbers of neurons in the Kohonen network.


We took experiments of cell cultures treated by eicosatetraynoic, linoleic and oleic acid. Number of experimetal data for eicosatetraynoic acid was 12, for linoleic acid was 40 and for oleic acid was 54. Experimental data we used for learning the Kohonen neural network. To test the algorithm, we reproduced relations between genes that are described in paper [3]. There table below contains genes at rows and columns with the cells containing the number of neurons in self organized map. By (*) we mark connections found in [3].
MT0082 MT0082 grpE pks16 fxsA MT2727 nei1 rpiB trxB grpE pks16 fxsA MT2727 nei1 rpiB trxB

* * * * * 7x5 *

5x3

* * * 9x7 * * * 5x3 -

References 1. Edgar, R., Domrachev, M., & Lash, A. E. (2002). Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic acids research, 30(1), 207-210. 2. Kohonen, T. (1990). The self-organizing map. Proceedings of the IEEE, 78(9), 1464-1480. 3. Sherrid, A. M., Rustad, T. R., Cangelosi, G. A., & Sherman, D. R. (2010). Characterization of a Clp protease gene regulator and the reaeration response in Mycobacterium tuberculosis. PLoS One, 5(7), e11622.