GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis

GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis Zefang Tang Boxi Kang Chenwei Li Tianxiang Chen Zemin Zhang

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis 2019 Nucleic Acids Research 10.1093/nar/gkz430 tang2019-gepia2-an-enhanced-web-server-for tang2019-gepia2-an-enhanced-web-server-for.txt, the project's staged copy d1b83dd4a0455f9c0dc9384876a8be655b20e675 cc-by genetics

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

0 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them gene- and isoform-level expression across TCGA tumour and GTEx normal samples Quotation withheld because this record is not cleared for redistribution. how to make expression, survival and signature analysis of TCGA/GTEx available at isoform and cancer-subtype resolution Quotation withheld because this record is not cleared for redistribution. a web server extending expression quantification from the gene to the transcript level and to 84 cancer subtypes, with user-uploaded data comparison and a python API Quotation withheld because this record is not cleared for redistribution. method Quotation withheld because this record is not cleared for redistribution. prognostic associations that appear only at cancer-subtype resolution Quotation withheld because this record is not cleared for redistribution. system-or-tool-building system-or-tool-building Quotation withheld because this record is not cleared for redistribution. alternative splicing shifts which isoform of a gene predominates, and the shift can be specific to a cancer type isoform switching Quotation withheld because this record is not cleared for redistribution. a user's uploaded matrix and the server-side target set are quantile-normalised against each other before any differential test, so the comparison is between normalised distributions rather than raw counts cross-dataset normalisation Quotation withheld because this record is not cleared for redistribution. the Cancer-subtype Classifier is described as predicting accurately, and the article reports no accuracy, no held-out split and no confusion matrix anywhere Quotation withheld because this record is not cleared for redistribution. the two classifier modes trade sensitivity against coverage, and the paper states the trade-off without quantifying either side Quotation withheld because this record is not cleared for redistribution. prognostic gene lists in the prior literature were derived at cancer-type level, so a subtype-specific effect is invisible in them Quotation withheld because this record is not cleared for redistribution. gene signature quantification, taken from single-cell sequencing practice into bulk TCGA/GTEx analysis Quotation withheld because this record is not cleared for redistribution. any accuracy, precision, recall or held-out validation figure for the Cancer-subtype Classifier eliminates: a classifier that is confidently wrong - no accuracy, precision, recall or held-out figure is given for a tool the sentence says predicts cancer subtype 'accurately' · outcome: absent-declared-unreported Quotation withheld because this record is not cleared for redistribution. the same MAPK6 survival test run at whole-breast-cancer level alongside the Lumina A subtype level eliminates: a survival signal that exists only in one subtype being reported as a property of the cancer type · outcome: held Quotation withheld because this record is not cleared for redistribution. signature score signature score Quotation withheld because this record is not cleared for redistribution. isoform usage isoform usage Quotation withheld because this record is not cleared for redistribution. survival map survival map Quotation withheld because this record is not cleared for redistribution. hazard ratios hazard ratios Quotation withheld because this record is not cleared for redistribution. cancer subtype cancer subtype Quotation withheld because this record is not cleared for redistribution. MSI MSI Quotation withheld because this record is not cleared for redistribution. MSS MSS Quotation withheld because this record is not cleared for redistribution. Quantile-normalization Quantile-normalization Quotation withheld because this record is not cleared for redistribution. Pfam Pfam Quotation withheld because this record is not cleared for redistribution. triple-negative breast cancer triple-negative breast cancer Quotation withheld because this record is not cleared for redistribution. MAPK6 MAPK6 Quotation withheld because this record is not cleared for redistribution. TPM TPM Quotation withheld because this record is not cleared for redistribution. log2(TPM + 1) log2(TPM + 1) Quotation withheld because this record is not cleared for redistribution. cancer type cancer type Quotation withheld because this record is not cleared for redistribution. breast cancer subtypes breast cancer subtypes Quotation withheld because this record is not cleared for redistribution. GEPIA2 GEPIA2 Quotation withheld because this record is not cleared for redistribution. GenomeSpace GenomeSpace Quotation withheld because this record is not cleared for redistribution. isoform protein domain structure plots isoform protein domain structure plots Quotation withheld because this record is not cleared for redistribution. GEPIA (Gene Expression Profiling Interactive Analysis) web server GEPIA (Gene Expression Profiling Interactive Analysis) web server Quotation withheld because this record is not cleared for redistribution. tumor and normal samples tumor and normal samples Quotation withheld because this record is not cleared for redistribution. TCGA TCGA Quotation withheld because this record is not cleared for redistribution. GTEx databases GTEx databases Quotation withheld because this record is not cleared for redistribution. updated and enhanced version updated and enhanced version Quotation withheld because this record is not cleared for redistribution. higher resolution higher resolution Quotation withheld because this record is not cleared for redistribution. 198 619 isoforms 198 619 isoforms Quotation withheld because this record is not cleared for redistribution. 84 cancer subtypes 84 cancer subtypes Quotation withheld because this record is not cleared for redistribution. gene expression quantification gene expression quantification Quotation withheld because this record is not cleared for redistribution. gene level gene level Quotation withheld because this record is not cleared for redistribution. transcript level transcript level Quotation withheld because this record is not cleared for redistribution. gene signature quantification gene signature quantification Quotation withheld because this record is not cleared for redistribution. single-cell sequencing studies single-cell sequencing studies Quotation withheld because this record is not cleared for redistribution. customized analysis customized analysis Quotation withheld because this record is not cleared for redistribution. own RNA-seq data own RNA-seq data Quotation withheld because this record is not cleared for redistribution. API for batch process API for batch process Quotation withheld because this record is not cleared for redistribution. easy retrieval of the analysis results easy retrieval of the analysis results Quotation withheld because this record is not cleared for redistribution. data mining data mining Quotation withheld because this record is not cleared for redistribution. gene functions gene functions Quotation withheld because this record is not cleared for redistribution. Survival analysis Survival analysis Quotation withheld because this record is not cleared for redistribution. gene expression levels gene expression levels Quotation withheld because this record is not cleared for redistribution. prognostic outcomes prognostic outcomes Quotation withheld because this record is not cleared for redistribution. clinical relevance of a given gene clinical relevance of a given gene Quotation withheld because this record is not cleared for redistribution. differential expression analysis differential expression analysis Quotation withheld because this record is not cleared for redistribution. similar gene identification similar gene identification Quotation withheld because this record is not cleared for redistribution. potential biomarkers potential biomarkers Quotation withheld because this record is not cleared for redistribution. therapeutic targets therapeutic targets Quotation withheld because this record is not cleared for redistribution. alternative splicing alternative splicing Quotation withheld because this record is not cleared for redistribution. different isoform usages different isoform usages Quotation withheld because this record is not cleared for redistribution. UCSC Xena recomputed data UCSC Xena recomputed data Quotation withheld because this record is not cleared for redistribution. coding coding Quotation withheld because this record is not cleared for redistribution. non-coding transcripts non-coding transcripts Quotation withheld because this record is not cleared for redistribution. heterogeneous subtypes with distinct prognosis heterogeneous subtypes with distinct prognosis Quotation withheld because this record is not cleared for redistribution. single-cell sequencing single-cell sequencing Quotation withheld because this record is not cleared for redistribution. gene signatures gene signatures Quotation withheld because this record is not cleared for redistribution. composition of tumor microenvironment composition of tumor microenvironment Quotation withheld because this record is not cleared for redistribution. functional status of infiltrating immune content functional status of infiltrating immune content Quotation withheld because this record is not cleared for redistribution. curated list of signatures curated list of signatures Quotation withheld because this record is not cleared for redistribution. Expression Analysis Expression Analysis Quotation withheld because this record is not cleared for redistribution. Custom Data Analysis Custom Data Analysis Quotation withheld because this record is not cleared for redistribution. eight tabs eight tabs Quotation withheld because this record is not cleared for redistribution. two tabs two tabs Quotation withheld because this record is not cleared for redistribution. Cancer Subtype Classifier Cancer Subtype Classifier Quotation withheld because this record is not cleared for redistribution. Expression Comparison Expression Comparison Quotation withheld because this record is not cleared for redistribution. python package python package Quotation withheld because this record is not cleared for redistribution. survival analyses survival analyses Quotation withheld because this record is not cleared for redistribution. custom cancer types custom cancer types Quotation withheld because this record is not cleared for redistribution. heat map heat map Quotation withheld because this record is not cleared for redistribution. statistical significance statistical significance Quotation withheld because this record is not cleared for redistribution. context-dependent context-dependent Quotation withheld because this record is not cleared for redistribution. prognostic impact prognostic impact Quotation withheld because this record is not cleared for redistribution. users’ custom data users’ custom data Quotation withheld because this record is not cleared for redistribution. public cancer genomic data public cancer genomic data Quotation withheld because this record is not cleared for redistribution. other two additional normalization strategies other two additional normalization strategies Quotation withheld because this record is not cleared for redistribution. uploaded dataset uploaded dataset Quotation withheld because this record is not cleared for redistribution. target dataset on the server target dataset on the server Quotation withheld because this record is not cleared for redistribution. boxplot boxplot Quotation withheld because this record is not cleared for redistribution. gene expression gene expression Quotation withheld because this record is not cleared for redistribution. TCGA expression data TCGA expression data Quotation withheld because this record is not cleared for redistribution. probability of this sample belonging to each subtype probability of this sample belonging to each subtype Quotation withheld because this record is not cleared for redistribution. two modes two modes Quotation withheld because this record is not cleared for redistribution. more sensitive result more sensitive result Quotation withheld because this record is not cleared for redistribution. cancer with unknown origin cancer with unknown origin Quotation withheld because this record is not cleared for redistribution. isoform symbol isoform symbol Quotation withheld because this record is not cleared for redistribution. Ensembl ID Ensembl ID Quotation withheld because this record is not cleared for redistribution. differentially expressed isoforms differentially expressed isoforms Quotation withheld because this record is not cleared for redistribution. prognostic isoforms prognostic isoforms Quotation withheld because this record is not cleared for redistribution. correlation between two isoforms correlation between two isoforms Quotation withheld because this record is not cleared for redistribution. HTML5, JavaScript, and PHP HTML5, JavaScript, and PHP Quotation withheld because this record is not cleared for redistribution. R and Perl scripts R and Perl scripts Quotation withheld because this record is not cleared for redistribution. uniform pipeline uniform pipeline Quotation withheld because this record is not cleared for redistribution. MySQL relational database MySQL relational database Quotation withheld because this record is not cleared for redistribution. GEPIA2 python API GEPIA2 python API Quotation withheld because this record is not cleared for redistribution. multiple requests multiple requests Quotation withheld because this record is not cleared for redistribution. efficiency of users efficiency of users Quotation withheld because this record is not cleared for redistribution. 5 functionalities 5 functionalities Quotation withheld because this record is not cleared for redistribution. immune-cell clusters immune-cell clusters Quotation withheld because this record is not cleared for redistribution. pan-cancer manner pan-cancer manner Quotation withheld because this record is not cleared for redistribution. Th1-like signature Th1-like signature Quotation withheld because this record is not cleared for redistribution. stomach tumors stomach tumors Quotation withheld because this record is not cleared for redistribution. adjacent normal tissues adjacent normal tissues Quotation withheld because this record is not cleared for redistribution. isoform quantification isoform quantification Quotation withheld because this record is not cleared for redistribution. cancer subtype information cancer subtype information Quotation withheld because this record is not cleared for redistribution. isoform usage profiling isoform usage profiling Quotation withheld because this record is not cleared for redistribution. custom data analysis custom data analysis Quotation withheld because this record is not cleared for redistribution. big cancer genomics data big cancer genomics data Quotation withheld because this record is not cleared for redistribution. GEPIA2 → GEPIA (Gene Expression Profiling Interactive Analysis) web server GEPIA2 is the updated and enhanced version of GEPIA. Version-of is absent from the eight edge types. Quotation withheld because this record is not cleared for redistribution. gene signature quantification → single-cell sequencing studies GEPIA2's signature quantification is inspired by techniques from single-cell sequencing studies. Inspired-by is absent. Quotation withheld because this record is not cleared for redistribution. Cancer Subtype Classifier → TCGA expression data The classifier is trained on TCGA expression data. Trained-on is absent from the governed types. Quotation withheld because this record is not cleared for redistribution. HTML5, JavaScript, and PHP → GEPIA2 The website's interactive controls are implemented with HTML5, JavaScript and PHP. Implemented-with is absent. Quotation withheld because this record is not cleared for redistribution. R and Perl scripts → GEPIA2 GEPIA2 accesses and processes server-side data using R and Perl scripts. Implemented-with is absent. Quotation withheld because this record is not cleared for redistribution. GEPIA2 python API → 5 functionalities The Python package provides access to five named GEPIA2 functions. Provides-access-to is absent from the edge registry. Quotation withheld because this record is not cleared for redistribution. signature score = mean log2(TPM + 1) across a signature's genes Quotation withheld because this record is not cleared for redistribution. 'Stacked' = pool the selected subtypes, then compare the pool with normal samples Quotation withheld because this record is not cleared for redistribution. 'Separated' = compare each selected subtype with normal samples on its own Quotation withheld because this record is not cleared for redistribution. 'Comparing Across Subtypes' = compare the selected subtypes with each other, not with normal Quotation withheld because this record is not cleared for redistribution. isoform 'switch' = the isoform carrying most of a gene's expression differs by cancer type Quotation withheld because this record is not cleared for redistribution. GEPIA2 - enhanced GEPIA with transcript-level, subtype-level, uploaded-data, and batch-API analysis Quotation withheld because this record is not cleared for redistribution. Resolution shift - expression quantification moves from gene level to individual transcripts and cancer subtypes Quotation withheld because this record is not cleared for redistribution. Survival analysis - links expression level to prognostic outcome to assess clinical relevance Quotation withheld because this record is not cleared for redistribution. Gene signature - curated gene set used to interrogate tumor-microenvironment composition or infiltrating-immune functional state Quotation withheld because this record is not cleared for redistribution. Survival map - context-sensitive gene or isoform prognosis summarized across selected cancer types Quotation withheld because this record is not cleared for redistribution. Uploaded expression comparison - normalize a user matrix with a public target and return differential expression as a boxplot Quotation withheld because this record is not cleared for redistribution. Classifier modes - known cancer type gives sensitive subtype prediction; unrestricted mode addresses unknown origin Quotation withheld because this record is not cleared for redistribution. Isoform analysis - profile expression, differential status, prognosis, and pairwise correlation using transcript identifiers Quotation withheld because this record is not cleared for redistribution. Python API - batch requests and streamlined retrieval for five analysis functions Quotation withheld because this record is not cleared for redistribution. extends → GEPIA (Tang et al. 2017), the previous version of this server note: the whole paper is a version increment; the extension named first is gene level to transcript level Quotation withheld because this record is not cleared for redistribution. qualifies → prognostic gene lists identified at the level of a general cancer type note: not a contradiction - the paper accepts those lists and argues they are at the wrong resolution Quotation withheld because this record is not cleared for redistribution. signature score —operationalised-as→ log2(TPM + 1) signature score log2(TPM + 1) the relation itself is separately evidenced · strength of the relation: 0.95 · the relation stated: the signature score IS DEFINED AS the mean of this transformed value over a signature's genes Quotation withheld because this record is not cleared for redistribution. MSI —distinct-from→ MSS MSI MSS differs in: response to chemotherapy and immunotherapy · strength of the relation: 0.95 Quotation withheld because this record is not cleared for redistribution. triple-negative breast cancer —distinct-from→ breast cancer subtypes triple-negative breast cancer breast cancer subtypes differs in: five-year survival rate, about 15% lower · strength of the relation: 0.85 Quotation withheld because this record is not cleared for redistribution. cancer subtype —part-of→ cancer type cancer subtype cancer type the relation itself is separately evidenced · strength of the relation: 0.95 · the relation stated: a cancer is COMPRISED BY its subtypes, so a subtype sits inside a type Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span writes 'cancers are often comprised by heterogeneous subtypes', not 'cancer subtype' / 'cancer type'. Quotation withheld because this record is not cleared for redistribution. GEPIA2 —contrasted-with→ GenomeSpace GEPIA2 GenomeSpace strength of the relation: 0.95 · note: the paper's only named alternative, and it concedes the alternative can do the same analyses Quotation withheld because this record is not cleared for redistribution. Pfam —operationalised-as→ isoform protein domain structure plots Pfam isoform protein domain structure plots strength of the relation: 0.85 Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span says the plots are BASED ON Pfam predictions; operationalised-as is loose. Quotation withheld because this record is not cleared for redistribution. alternative splicing —causes→ different isoform usages alternative splicing different isoform usages the relation itself is separately evidenced · strength of the relation: 1 · note: The paper states that alternative splicing results in different isoform usages. Quotation withheld because this record is not cleared for redistribution. GEPIA (Gene Expression Profiling Interactive Analysis) web server —causes→ potential biomarkers GEPIA (Gene Expression Profiling Interactive Analysis) web server potential biomarkers the relation itself is separately evidenced · strength of the relation: 1 · note: GEPIA use led to identification of potential biomarkers. Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span writes 'the GEPIA web server', not 'GEPIA (Gene Expression Profiling Interactive Analysis) web server'. Quotation withheld because this record is not cleared for redistribution. GEPIA (Gene Expression Profiling Interactive Analysis) web server —causes→ therapeutic targets GEPIA (Gene Expression Profiling Interactive Analysis) web server therapeutic targets the relation itself is separately evidenced · strength of the relation: 1 · note: GEPIA use also led to identification of therapeutic targets. Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span writes 'the GEPIA web server', not 'GEPIA (Gene Expression Profiling Interactive Analysis) web server'. Quotation withheld because this record is not cleared for redistribution. GEPIA2 python API —causes→ efficiency of users GEPIA2 python API efficiency of users the relation itself is separately evidenced · strength of the relation: 1 · note: The API boosts efficiency through batch processing and streamlined retrieval. Quotation withheld because this record is not cleared for redistribution. stomach tumors —distinct-from→ adjacent normal tissues stomach tumors adjacent normal tissues the relation itself is separately evidenced · strength of the relation: 1 · note: The Th1-like signature is significantly higher in tumors than adjacent normal tissue. Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but the span says the Th1-like signature is higher in stomach tumors than in adjacent normal tissue; distinct-from is loose. Quotation withheld because this record is not cleared for redistribution. MAPK6 shows no association with breast cancer survival overall, and a significant unfavourable association within the Lumina A subtype Quotation withheld because this record is not cleared for redistribution. the Th1-like signature score is significantly higher in stomach tumours than in adjacent normal tissue Quotation withheld because this record is not cleared for redistribution. gene survival analyses are context-dependent, which is the stated reason for scanning many cancer types at once Quotation withheld because this record is not cleared for redistribution. signature score signature score Quotation withheld because this record is not cleared for redistribution. isoform usage isoform usage Quotation withheld because this record is not cleared for redistribution. Stacked Stacked Quotation withheld because this record is not cleared for redistribution. Separated Separated Quotation withheld because this record is not cleared for redistribution. Comparing Across Subtypes Comparing Across Subtypes Quotation withheld because this record is not cleared for redistribution. 198 619 isoforms — None Quotation withheld because this record is not cleared for redistribution. 84 cancer subtypes — None Quotation withheld because this record is not cleared for redistribution. about 15% lower five-year survival in triple-negative breast cancer — None Quotation withheld because this record is not cleared for redistribution. 414 citations of GEPIA by May 2019, per Google Scholar — None Quotation withheld because this record is not cleared for redistribution. about 400 000 analysis requests from about 110 000 distinct IP addresses — None Quotation withheld because this record is not cleared for redistribution. 5 functionalities in the python package — None Quotation withheld because this record is not cleared for redistribution. MySQL version 5.7.17 — None Quotation withheld because this record is not cleared for redistribution. log2(TPM + 1) log2(TPM + 1) Quotation withheld because this record is not cleared for redistribution. hazard ratio in log10 scale hazard ratio in log10 scale Quotation withheld because this record is not cleared for redistribution. P = 0.022 P = 0.022 Quotation withheld because this record is not cleared for redistribution. P = 0.15 P = 0.15 Quotation withheld because this record is not cleared for redistribution. Quantile-normalization Quantile-normalization Quotation withheld because this record is not cleared for redistribution. isoform expression level isoform expression level Quotation withheld because this record is not cleared for redistribution. signature score signature score Quotation withheld because this record is not cleared for redistribution. hazard ratio hazard ratio Quotation withheld because this record is not cleared for redistribution.

The full text of this article is not redistributable, so it is not included here. The evidence sentences are withheld with it; the tag layer and its grounding status remain. The article itself is at https://doi.org/10.1093/nar/gkz430.