Description

VarChat is an open platform that leverages the power of generative artificial intelligence to support the genomic variant interpretation process by searching the available scientific literature for each variant and condensing it into a brief yet informative text. Each query quickly scans the latest scientific literature to provide up-to-date variant information.

VarChat is a generative AI-based system and each answer is generated live, so you may obtain slightly different answers at each iteration. A literature search will be performed and the total number of identified publications will be shown. Only a subset of them will be reported and used to generate your answer.

If you would like to stay updated on the latest developments, you may register for updates on the VarChat website. For data questions, VarChat can be contacted at [email protected].

Display Conventions and Configuration

Genomic locations of variants are labeled with the nucleotide change. Mousing over the items will show how many papers the variant was observed in, its gene, its HGVS nomenclature, and dbSNP rsID. Clicking on any item will provide a link directly to VarChat with additional information.

The items are colored based on the amount of literature support as described on the table below:

Methods

VarChat software is powered by enGenome.
enGenome, an accredited spin-off from the University of Pavia founded in 2016, combines bioinformatics, biotechnology, and software development expertise to enhance genetic disease diagnosis and treatment through advanced AI and bioinformatics tools, supported by a multidisciplinary team of engineers, biotechnologists, and developers.

For every queried variant, VarChat produces concise and coherent summaries through an LLM model. Relevant references are identified through a modified BM25 ranking algorithm. More weight is given to papers that cite the variant in the abstract and were published in the last two years, while papers that report the variant only in the supplementary are penalized.

Data access

The raw data can be explored interactively with the Table Browser or the Data Integrator. The data can be accessed from scripts through our API, the track name is "varChat".

For automated download and analysis, the genome annotation is stored in a bigBed file that can be downloaded from our download server. The file for this track is called varChat.bb. Individual regions or the whole genome annotation can be obtained using our tool bigBedToBed which can be compiled from the source code or downloaded as a precompiled binary for your system.

Instructions for downloading source code and binaries can be found here. The tool can also be used to obtain only features within a given range, e.g.

bigBedToBed http://hgdownload.soe.ucsc.edu/gbdb/hg38/bbi/varChat.bb -chrom=chr21 -start=0 -end=10000000 stdout

References

De Paoli F, Berardelli S, Limongelli I, Rizzo E, Zucca S. VarChat: the generative AI assistant for the interpretation of human genomic variations. Bioinformatics. 2024Mar29;40(4). PMID: 38579245; PMC: PMC11055464