PROGgeneV2: enhancements on the existing database.

Goswami CP, Nakshatri H. · 2014

Read PROGgeneV2: enhancements on the existing database. by Goswami CP, Nakshatri H. (2014) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc4300843.txt.

Introduction

Chirayu Pankaj Goswami and Harikrishna Nakshatri describe PROGgeneV2, an updated web database that links gene activity in tumors to patient survival. Their paper appeared in BMC Cancer in 2014. The first version let users ask whether a gene's expression level in a cancer tracked with better or worse outcomes, drawn from public studies.

The update nearly doubles the catalog of studies and widens what users can test. Survival can now be checked for single genes, multi gene signatures, and ratios between two genes, and models can adjust for patient factors such as age. Published gene signatures are included ready made. The tool serves cancer researchers who want a fast prognosis check before deeper work.

Catalog record

CREATORGoswami CP, Nakshatri H.
YEAR2014
COLLECTIONPAPERS
FORMATTXT
FILE SIZE19 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC4300843

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

PROGgeneV2: enhancements on the existing database By Chirayu Pankaj Goswami, Harikrishna Nakshatri Published in: BMC Cancer DOI: 10.1186/1471-2407-14-970 ABSTRACT Background We recently published PROGgene, a tool that can be used to study prognostic implications of genes in various cancers. The first version of the tool had several areas for improvement. In this paper we present some major enhancements we have made on the existing tool in the new version, PROGgeneV2. Results In PROGgeneV2, we have made several modifications to enhance survival analysis capability of the tool. First, we have increased the repository of public studies catalogued in our tool by almost two folds. We have also added additional functionalities to perform survival analysis in a variety of new ways. Survival analysis can now be performed on a) single genes b) multiple genes as a signature, c) ratio of expression of two genes, and d) curated/published gene signatures in new version. Users can now also adjust the survival analysis models for available covariates. Users can study prognostic implications of entire gene signatures in different cancer types, which are searchable by keywords. Also, unique to our tool, in the new version, users will be able to upload and use their own datasets to perform survival analysis on genes of interest. Conclusions We believe, like its predecessor, PROGGeneV2 will continue to be useful for the scientific community for formulating research hypotheses and designing mechanistic studies. With added datasets PROGgeneV2 is the most comprehensive survival analysis tool available. PROGgeneV2 is available at http://www.compbio.iupui.edu/proggene. Electronic supplementary material The online version of this article (doi:10.1186/1471-2407-14-970) contains supplementary material, which is available to authorized users. BACKGROUND Survival analysis using gene expression to derive predictive gene signatures is a commonly used feature in research studies employing high throughput genomic data. Gene signatures predictive of overall, relapse free or metastasis free survival are popular and several such signatures are published periodically and the data submitted to public repositories. Data from such studies which is available on the public domain can be leveraged to identify prognostic markers in different cancer types. We recently published ‘PROGgene’ [1], a tool that allows researchers to be able to use publicly available data to study prognostic implications of genes of interest in multiple cancers. This tool can be used to conduct on the fly survival analysis and create survival plots (Kaplan Meier, KM plots) based on gene expression of user input genes in user selected datasets from multiple cancers. Since its publication, PROGgene has received encouraging response. In the first version of the tool, we catalogued data from sixty-four studies from eighteen major cancer types. The data was primarily downloaded from Gene Expression Omnibus and The Cancer Genome Atlas. Version 1 of the tool had several areas for improvement. In this paper, we present the second release of PROGgene, PROGgeneV2, which has several enhancements over the previous version. PROGgeneV2 will be available freely for research and non-commercial use at the same URL as its predecessor, http://www.compbio.iupui.edu/proggene. In the new version, we have added additional datasets for existing and new cancer types. We have also added more functionality to the tool which will allow researchers to perform multifaceted survival analysis. IMPLEMENTATION New and additional datasets The data in PROGgeneV2 are downloaded from published studies where data has been deposited to public repositories. The studies included have acquired proper authorization to release data based on human subjects. The first release of PROGgene catalogued sixty four cohorts from eighteen cancer types. In the new version, we have expanded the repository of the tool more than 2 folds by adding additional datasets. The tool now contains data from 134 cohorts from 21 cancer types. Figure 1 shows the number of datasets available in PROGgeneV1 and V2. Ten new datasets have been added for Lung and Colon cancers each in the new version. Datasets for Esophageal, Cervical and Eye cancers have been included for the first time. Figure 2 shows the number of samples in datasets that are included in the new version. Additional file 1: Table S1 lists the new datasets available in PROGgeneV2. For survival variables and covariates available for the new datasets, please refer to Additional file 1: Table S2.Figure 1 Comparison of number of datasets available in PROGgeneV1 and V2. Figure 2 Number of samples in datasets that are included in PROGgeneV2. New features In the first version of PROGgene, users were not able to study prognostic implications of ratio of expression of two genes. In the new version of PROGgene, we have added capability to study prognostic implications of a gene expression ratio. Users can input two genes in the home page and create survival plots for the gene expression ratio in cohorts of interest. In the new version, we have also added capability where users can create survival plots for published/curated gene signatures. We have included approximately ten thousand published/curated gene signatures from Molecular signature database [2] to our repository. These signatures include those from curated databases such as Gene Ontology [3], KEGG [4], Biocarta and Reactome [5] and from published studies. Users can do a keyword search on the gene signatures and the application retrieves genes included in the gene signature as provided by the Molecular Signature Database. When using gene signatures, only a combined plot using mean of expression of all genes in the signature is created for the entire signature. Also added in the new version is the functionality to adjust survival models for covariates. Covariate data such as clinical variables is submitted along with gene expression data in public repositories. We have included the covariates available for datasets in the new version of the application. Survival models can be adjusted for multiple covariates. In PROGgeneV1, users were able to only divide the cohort into subgroups based on covariates and study survival implications of these genes in the subgroups, eg Stages I, II and III etc. This feature still remains in the new version. Another important improvement made in the new version is capability to upload custom datasets. Users can now upload their own datasets to PROGGeneV2. The instructions to upload are provided on the application website. Upon upload of custom datasets, users can use custom datasets just like any default dataset on PROGGeneV2 to conduct survival analysis on genes in their dataset. Also, the prognostic plots in the new version are accompanied with summary statistics, such as no of events, median survival times for survival categories, confidence intervals etc. Lastly, in the new version we have also provided option for bifurcating the gene expression variable for proportional hazards analysis using mean, median, 25th percentile or 75th percentile of gene expression. Workflow Similar to its predecessor, PROGgeneV2 is written in PHP5 with a MySQL database backend which stores gene expression data, covariates data and metadata for catalogued studies in form of relational database tables. Survival analysis is done using backend R script which employs R library ‘survival’ to perform Cox proportional hazards analysis (function ‘coxph’) and to plot prognostic plots (function ‘survfit’). The functioning of website is very similar to the previous version. General workflow of the application is presented in Figure 3. In first step users input the genes of interest on the ‘home page’. Users also select analysis

Read the complete text