- interfaces and the possibility to upload custom datasets.. - Full list of author information is available at the end of the article. - 1): i) GMIEC-AN is used for the analysis of a minimum of two datasets. - ii) GMIEC-results parses the output of GMIEC-AN, performs the automatic selection of the modules and outputs the results. - iii) GMIEC-VIS, is a tool that helps to explore the results of GMIEC-AN.. - In fact, the user can perform one analysis consid- ering gene-expression, copy-number variation, methyla- tion and mutation data and using only two or more of this kind of data. - The main output of GMIEC is a tab-delimitated file (green box), in which the rows are the samples and columns the modules. - GMIEC-results is a sub-application that performs the automatic selection of the modules in each patient using the output from the analysis with GMIEC-AN. - This sim- plified version of the output file can be also uploaded on GMIEC-results (blue arrow) to visualize the levels of the gene-expression, methylation and copy number vari- ation, mutations (or the omics data provided) of the genes inside one module of a specific patient. - GMIEC- VIS provides functionalities to explore the results of GMIEC-AN using a dynamic interface with heatmaps, charts and tables.. - GMIEC-AN implements two strategies of analysis (called M1 and M2), (see section “Computational Methods”. - The analytical pro- cedure implemented in M1 in the first step, selects only the profiles of the genomics data of a given set of genes G (see section “GMIEC framework of analysis” Add- itional File 1 and Fig. - In the first step, the user uploads the omics datasets using a gui-interface. - The main application of GMIEC, GMIEC-AN implements two methods of analysis (M1, M2). - GMIEC-AN allows to analyze two (orange) or more datasets. - When the user provides a file with the genomic coordinates (e.g. - The output of GMIEC-AN is a tab delimited file containing the genes and drugs modules identified during the analysis. - Otherwise, the user can upload the output of GMIEC-AN on GMIEC-results. - by the user. - For example, the user can upload a list of genes of interest. - gene- expression, copy-number) for the current i. - For each patient, the number of rows and columns of this matrix correspond respectively to the number of genes in G and the num- ber of datasets uploaded by the user. - gene-expression, copy-number alteration). - The proximity matrix is a symmetric matrix in which each value is the proportion of times that each pair of genes resides on the same terminal node of the tree. - Then, the matrix is used as input for a k-means method to identify clusters of genes (step 5, see section “GMIEC framework of analysis” Additional file 1, Fig. - S1) with a pre-defined number of clusters chosen by the user. - GMIEC-AN can also determine the optimal number of clusters using silhouette analysis. - Therefore, each GM is merged with a file containing the association be- tween the genes and the drugs (step 6, see section. - “GMIEC framework of analysis” Additional file 1, Fig.. - The user can create a custom file containing the as- sociations between specific genes and drugs, or down- load gene/drugs datasets available online [9]. - S1) for the genomic alterations and the fraction of drugs identified in a module. - copy-number variation). - For all genes (g) of a patient (i) a matrix M i = (g x r), in which n are the genes and r the rules, is generated.. - The logic rules imple- mented are the same if the user upload more or two datasets. - The number of clusters to use in k-modes is defined by the user.. - Several tools and databases have been developed during the last years with the aim to facilitate the integration and analysis of multi-omics data in the field of the per- sonalized medicine. - Here we report a description of these resources and the comparisons with GMIEC (see Table 1). - Users can query the database providing the name of a gene and the type of a cancer. - The database implements an interface to com- pute the survival probabilities of the patients between cancers. - The user can query the database using the name of a gene or drug.. - DriverDBv3 and PanDrugs allows the user to use the da- tabases with multiple options. - Other two databases PreMedKB [6] and OncoKB [13] does not allow the user to upload their datasets. - Notably, Cancer Variant Explorer [19] is a shiny and R-package that per- form the prioritization of the variants to identify mecha- nisms that drives the resistance and the druggability of a treatment. - omics data: gene-expression, copy-number, epigen- etics and mutations data in the form of tab- delimited files. - In the case the genetic variants data, the user might create a file with only the mutations of interest (e.g. - The user can create a custom file selecting only the drugs and genes that are important for the disease of interest or download them from our repository.. - clinical data: an optional tab delimited file with clinical variables to include in the analysis.. - The list of genes can be directly uploaded by the user. - Otherwise if the user provides a file with genomic coordinates, the list of genes is obtained through an annotation step.. - The principal output of GMIEC are table delimited files that can be easily opened with Microsoft Excel or Libre/Open Office. - The out- put reports also columns containing the scores (see sec- tion “Computation of the scores” Additional file 1) that are a quantification of the level of genomic alterations of the genes in GM, and the fraction of drugs linked to. - GMIEC-VIS allows to visualize, for each patient, tables with statistics related to the gene modules, and the lists of drugs and genes identified. - GMIEC-results is another sub- application that allows to parse the output of GMIEC and automatically select a module for patient. - GMIEC- results can be also used to visualize the levels the omics data for the genes inside each module.. - The user can run the analysis (Fig. - 2a) and the “tabs” available:. - GMIEC-AN: GMIEC-Analysis (GMIEC-AN) is the principle module of GMIEC and allows to analyze multiple datasets. - The user can easily upload the data using dedicated fields and windows (Fig. - These windows allow to select the type of analysis (M1 or M2) and options (see section “ Options to run GMIEC- AN”, Additional File 1) to employ.. - Visualization (GMIEC-VIS): is another interface that allows to upload the results of GMIEC-AN and explore the data. - allows the visualization of heatmaps containing the values of drugs, genes, sad or S-score values computed by GMIEC-AN. - These heatmaps are interactive, the user can zoom in or out to explore the results (Fig. - The second section of GMIEC-VIS allows to inspect Table 1 Comparison between GMIEC and others bioinformatic tools available in precision medicine. - Comparison the bioinformatics tools and databases that can be used in the context of precision medicine. - The column “ name resource ” contains the name of the software/database. - the results of GMIEC at the level of single patient. - The user can select a given patient to visualize a table that contains several statistics for each module (e.g. - 2c), that facilitates the visualization of the modules, from the simplified ver- sion of the GMIEC output table obtained from GMIE C-RES. - This sub-application contains an interface simi- lar to those one of GMIEC and the user can upload the result of GMIEC, the omics data and visualize for each patient the gene-expression, copy-number, methylation levels, the mutational status and the number of the drugs for the genes identified in the module (Fig. - Results (GMIEC-RES): The output obtained from GMIEC-AN can be uploaded into GMIEC-results (see section “GMIEC-results” Additional File 1 for a complete description of GMIEC-RES). - This application allows i) the automatic selection of the modules from each patient. - ii) the parsing of GMIEC-AN output in a. - iii) visualize the results of GMIEC-AN.. - The concept underlying GMIEC-results is selecting group of the genes in one patient that are the most over-activated (e.g. - Although this approach is a simpli- fication, GMIEC-results it is a straightforward approach to support the users with few computational skills to parse the output of GMIEC-AN and easily identify the modules that could be biologically more relevant. - More expert users can use the original output file created by GMIEC-AN to select the modules using custom ap- proaches. - The output of GMIEC-results is a tab delim- ited file with the rows corresponding to the patients and four columns with the patient id, the score of the modules, the genes and the drugs inside each module.. - a The main interface of GMIEC. - b GMIEC-AN contains fields that support the users to upload the data. - The first one ‘ Summary Heatmaps GMIEC ’ allows the visualization of heatmaps containing the values of drugs, genes, sad or S-score ss computed by GMIEC-AN. - The second section of GMIEC-VIS allows to inspect the results of GMIEC at the level of single patient with tables. - The main purpose of this analysis (see “Analysis of patients with prostate cancer” Additional File 1, for de- tails of analysis) was the identification of active onco- genic modules (AOMs) associated with drugs (AOMDs) at the level of single patient. - When we explored the results of the patients with AOMDs (see “Analysis of patients with prostate cancer”. - Additional file 1, Fig. - We also selected the inactive oncogenic modules (IOMDs) associated with drugs (see “Analysis of patients with prostate cancer” Additional file 1, Fig. - “Analysis of patients with prostate cancer” Additional file 1, Fig. - For example, (see “Analysis of patients with prostate cancer” Additional file 1, Fig. - S2B) shows a patient with an up-regulation of the gene ERCC6, which is involved in DNA repair, associated with the molecule cisplatin in a specific patient. - Instead, (see “Analysis of patients with prostate cancer” Additional file 1, Fig. - These modules might be exploited in future to target the DDR pathway using a personalized approach based on the knowledge of the genetic status of the single patients. - The user can also explore the modules containing specifics genes (see section “Analysis of modules containing a specific gene” Additional file 1 and Fig. - The interface of GMIEC allows the user, with no advanced programming skills, to obtain an output that can be queried in a very easy way. - GMIEC-AN implements multiple options for the analysis and return as output a simple tab-delimited.. - Then, this output can be uploaded into two other sub- applications (GMIEC-results and GMIEC-VIS) to facili- tate the detection of gene modules and the exploration of the data. - can be useful in the context of the data integration, data mining and data-exploration at the level of single pa- tient. - Additional file 1. - This document contains an extended description of the methods implemented in GMIE C, the supplementary results and the supporting Figs. - GMIEC-AN: GMIEC-analysis. - The prostate cancer data [24] used to test the functionalities of GMIEC were downloaded from www.cbioportal.org [25, 26]. - DGIdb 3.0: a redesign and expansion of the drug-gene interaction database.
Xem thử không khả dụng, vui lòng xem tại trang nguồn hoặc xem
Tóm tắt