In order to focus effort on the most useful aspects of SFARI Gene to the autism research community, several changes are being made to the database. Given the regular citation of the human gene module in the primary literature, curation of the autism risk gene list will continue as before, using the current annotation model. Other modules will change as follows:
In order to simplify the gene scoring system we have decided to reduce the number of categories from seven to four. The ‘S’ (syndromic) category will remain, to include genes associated with medical syndromes that are frequently accompanied by autism diagnoses. There will then be three categories for genes associated with ‘idiopathic’ autism. The new category 1 (high confidence) genes will be taken from the SPARK gene list, which are genes for which there is sufficient evidence in the literature that they are deemed ‘returnable’ results to families in a clinical setting. In practice this means three de novo likely gene-disrupting mutations in individuals with an ASD diagnosis. All of these are in either category 1 or 2 under the previous scoring system. The new category 2 (strong candidate) genes will be drawn from the previous category 3, and the new category 3 (suggestive evidence) genes will be drawn from the previous category 4. The previous categories 5 and 6 will be eliminated. As always, we encourage users of SFARI Gene list to carry out analyses with individual categories, as well as with the entire list.
The scope of annotation for the CNV module will be scaled back by focusing on curation of recurrent CNVs associated with ASD (e.g. as defined in supplementary table 11 of Brandler et al., Science 360, 327-331 (2018). MindSpec will also curate single-gene CNVs with statistical significance and replication across studies (e.g. exonic CNVs overlapping NRXN1 at 2q16.3). Rare single instances of CNVs will no longer be annotated. The CNV scrubber will be populated by CNVs called from analyses of the Simons Simplex Collection (Sanders et al., Neuron 87, 1215-
1233 (2015), Werling et al., Nat. Genet. 50, 727-736 (2018), Collins et al., Genome Biol. 18, 36 (2017), and Coe et al., Nat. Genet. 46, 1063-1071 (2014).
Animal Model module
The scope of the Animal Model module will be substantially scaled back by focusing on curation of only high-impact mouse models of ASD identified based on stringent selection criteria.
Protein Interaction (PIN) module
The PIN module will be no longer be hosted on the current SFARI Gene site.