SelTarbase, main

NEWS
release 201006.25, june 2010

SelTarbase TOOLS (→ Services)
· MNR_ensembl v2.1: [extended] search for human (mouse) coding, untranslated, non-coding, and intronic mononucleotide repeats (based on Ensembl rel. 57_37b) including stratification by subcellular localization database LOCATE's information
· Predicted targets: genes with significantly altered mutation frequencies within MNRs compared to the background
· Submit data: upload your own new data and let the regression analysis recalculated including these data
· Register for free: use SelTarbase Tools and keep informed about recent changes
· Cancer Cell Line Infos: get detail information on human (MSI) cancer cell lines.

ABOUT SelTarbase
SelTarbase is a comprehensive mononucleotide repeat (MNR) mutation database. The primary data are derived from investigations of human microsatellite high-unstable (MSI-H) tumors of different organs. SelTarbase provides newest information of a very large, growing number of genes respectively the contained mononucleotide repeats. Additionally an up to date tissue specific regression analysis helps to decide which mutation frequencies seem to be elevated or reduced and could help focus direct investigation to promising candidate genes of MSI-H tumorigenesis. Furthermore, SelTarbase allows for upload of new (anonymized) data and recalculation of a regression analysis including these new data endowing the user with new aspects of his own research results. NAR 2010 Database Issue

ABOUT SelTarbase

SelTarbase is a comprehensive mononucleotide repeat (MNR) mutation database. The primary data are derived from investigations of human microsatellite high-unstable (MSI-H) tumors of different organs.

SelTarbase provides newest information of a very large, growing number of genes respectively the contained mononucleotide repeats. Additionally an up to date tissue specific regression analysis helps to decide which mutation frequencies seem to be elevated or reduced and could help focus direct investigation to promising candidate genes of MSI-H tumorigenesis. Furthermore, SelTarbase allows for upload of new (anonymized) data and recalculation of a regression analysis including these new data endowing the user with new aspects of his own research results.

NAR 2010 Database Issue

BACKGROUND
Deficient DNA mismatch repair (MMR deficiency) results in deletions or insertions in small repetitive DNA elements consisting of one, two or more nucleotides as single units, known as microsatellites. Microsatellite instability (MSI) occurs in more than 90% of human tumors in patients suffering from hereditary non-polyposis colorectal cancer (HNPCC/ Lynch syndrome, OMIM #120435) but also arises in sporadic carcinomas of the colon, endometrium, and stomach albeit at lower frequency (up to about 15%). The mutability of microsatellites mainly depends on their length and additionally on other biochemical and biological attributes, whose impact is not finally determined. Most of all genes respectively repeat tracts without physiologic relevance will show a similar mutation rate. However, some of the manifest mutations provide a positive or negative selection impact to affected cell clones leading to increased or reduced mutation frequencies. Therefore, the observed mutation rate may vary in a wide range. We have proposed a statistical model based on sigmoid regression analysis aiming at the identification of relevant genes of MSI driven carcinogenesis by their mutation frequency in regard to the repeat tract length. Extensive literature review leads to inclusion of datasets regarding a specific mononucleotide repeat tract (MNR) in the human genome as well as the number of analyzed MSI-H tumors of a certain tissue type (e. g. colon, stomach, and endometrium) and the number of tumors showing mutations within this MNR. Mutational data from the literature are collected and stored in a MySQL database. Tissue specific regression analyses are performed with R and nls2. All steps from database query and R calculation to complete web page presentation is done by a number of perl scripts.

BACKGROUND

Deficient DNA mismatch repair (MMR deficiency) results in deletions or insertions in small repetitive DNA elements consisting of one, two or more nucleotides as single units, known as microsatellites. Microsatellite instability (MSI) occurs in more than 90% of human tumors in patients suffering from hereditary non-polyposis colorectal cancer (HNPCC/ Lynch syndrome, OMIM #120435) but also arises in sporadic carcinomas of the colon, endometrium, and stomach albeit at lower frequency (up to about 15%). The mutability of microsatellites mainly depends on their length and additionally on other biochemical and biological attributes, whose impact is not finally determined. Most of all genes respectively repeat tracts without physiologic relevance will show a similar mutation rate. However, some of the manifest mutations provide a positive or negative selection impact to affected cell clones leading to increased or reduced mutation frequencies. Therefore, the observed mutation rate may vary in a wide range.

We have proposed a statistical model based on sigmoid regression analysis aiming at the identification of relevant genes of MSI driven carcinogenesis by their mutation frequency in regard to the repeat tract length. Extensive literature review leads to inclusion of datasets regarding a specific mononucleotide repeat tract (MNR) in the human genome as well as the number of analyzed MSI-H tumors of a certain tissue type (e. g. colon, stomach, and endometrium) and the number of tumors showing mutations within this MNR.

Mutational data from the literature are collected and stored in a MySQL database. Tissue specific regression analyses are performed with R and nls2. All steps from database query and R calculation to complete web page presentation is done by a number of perl scripts.

SUMMARY STATUS

The following table summarizes the contents of the selected and the previous release.

	release last	release latest
date	201005	201006

references analyzed	722	729
references included	540	546
colon	344	347
stomach	135	137
endometrium	82	85
colonculture	89	89
Genes analyzed	568	574
Genes included	529	535
colon	454	460
stomach	133	139
endometrium	81	81
colonculture	296	296
MNRs analyzed	4293	4326
MNRs included	3516 (2937 c., 579 nc.)	3549 (2953 c., 596 nc.)
colon	1909 (1506 c., 403 nc.)	1942 (1522 c., 420 nc.)
stomach	735 (707 c., 28 nc.)	744 (716 c., 28 nc.)
endometrium	795 (767 c., 28 nc.)	808 (780 c., 28 nc.)
colonculture	2071 (1826 c., 245 nc.)	2071 (1826 c., 245 nc.)
observations included
colon	151662	152445
stomach	30943	31181
endometrium	24912	25628
colonculture	25567	25567