The MTB portal is a clinical decision support system and data sharing infrastructure used across several comprehensive cancer centers connected by the Cancer Core Europe. Here, we provide an open lightweight version of its analytical framework that performs a general interpretation of the functional and predictive relevance of the gene variants uploaded by the user (see the FAQ page for more details).
The analyses provided by the public version of the MTB portal are limited to a general classification of the functional and predictive relevance of the variants uploaded by the user based on the combination of multiple knowledge sources, bioinformatics tools and data bases. Note that this public resource does not include variant interpretation nuances or ad-hoc clinical flags that are relevant for the clinical decision making, and thus this resource is intended for research purposes only (see the FAQ page for more details).
Results of the MTBP analyses are structured in HTML reports where variants are classified according to their functional relevance, and those described as cancer biomarkers are ranked according to their level of actionability. These documents provide detailed variant annotation, evidence supporting each given assertion and access to original data sources through interactive elements. This is aimed to empower in-depth review of available information.
The MTBP uses a comprehensive set of expert curated databases collecting clinical and preclinical knowledge associated to genomic events (such as their role in disease pathogenicity or anti-cancer drugs response). Assertions supported by lower quality data (e.g. not well-powered studies) and/or without consensus are not used to classify the variants (but provided as additional descriptives).
The MTBP uses public knowledgebases developed by international efforts that (a) formalize the variant information by using pre-defined criteria; (b) are based on published biomedical literature (including conference communications for most recent reports); (c) are committed to periodical updates; and (d) are open for the use and review of the community (see individual resources for licensing details).
Some bona fide assumptions are used to estimate the relevance of particular events without reported effects. These assumptions are based on stringent criteria, as the recommendations to consider mutations of null consequence type as loss-of-function events developed by the study of Mendelian gene variants.
Computational estimations are used as the lowest level of evidence to estimate the variants effect if no other information is available. Only results supported by robust statistical methods are employed according to our own benchmarking.
The portal is developed under the umbrella of Cancer Core Europe, a network of seven leading European cancer centers sharing knowledge, data and technology to implement new precision oncology strategies.
The public website of the MTBP provides a lightweight version of the analytical pipeline used by CCE initiatives, which supports a general interpretation of gene variants for research purposes.