Under construction.
Requirements for a VarImpact installation are the following.
Third-party software
- GNAT for the detection of gene mentions in text and mapping to NCBI Entrez Gene
- SETH to recognize genetic variants in text
- note that instead of GNAT and SETH, any NER tool for genes and mutations will work, as long as it can tag the recognized entities inline in XML documents, and map genes/proteins to identifiers such as official symbol, UniProt IDs, NCBI IDs; and mutations to HGVS nomenclature
- by default, VarImpact works with MutationFinder
Other hardware/software
- a local MySQL instance (any other RDBMS should work in theory): required by GNAT; needed for VarImpact only if you want to store the final results in an RDBMS
- an inhouse copy of PubMed/MEDLINE documents (XML)
- this could be limited to abstracts that you consider relevant
- VarImpact can be adapted to use non-MedlineCitationSet XML
Continue to [Usage] instructions.