Opportunities and Challenges for the Matchmaker Exchange
With greater than 14,000 unique genes connected across the MME, most matchmaking submissions result in at least one match (Figure 3 ) and each of these matches requires review that will need additional information to inform the matching parties as to whether or not the match represents a valid novel disease-gene relationship discovery. A matchmaking outcome analysis of 194 genes submitted to MME by the Care4Rare Canada Consortium over a 2-year period highlighted the effectiveness of MME in establishing collaborations for novel candidate genes with a 15% success rate (Osmond, Hartley, Dyment, et al., 2022) . The 194 genes entered by Care4Rare Canada resulted in over 1,500 matches returned by MME, with 93% receiving at least one match. Each of these needed to be evaluated, consolidated, and, in the majority of cases, have emails sent to determine the potential significance of the match; this is a laborious process that requires dedicated time and resources. Likewise, seqr users have supported manual communication on >6,500 matches (PRODUCTION: REFERENCE APPEARS IN THE SAME SPECIAL ISSUE (Pais et al., 2022) . One of the largest contributing factors to the post-match workload is the lack of phenotypic and inheritance information associated with the majority of cases in GeneMatcher, the largest node of the MME (Osmond, Hartley, Dyment, et al., 2022) . To date, only 2.5% of the over 63,000 submissions to GeneMatcher have included phenotype. However, the majority of records in all other primary MME nodes do include this information. The discovery community needs to carefully weigh the pros and cons of a low barrier to entry into MME with the post-match workload.
Moving forward, the MME has the opportunity to improve the efficiency of the matchmaking (Table 2 ). Inclusion of additional information (phenotype, variant details, zygosity) along with the gene will facilitate the ruling out of matches without the email back-and-forth that is difficult to scale. Osmond and colleagues (2022) showed that ~50% of matches could be ruled out when this type of data was provided by the node at the time of match notification. It is important for users to note that inclusion of high level phenotypic data typically does not require additional consent (Dyke et al., 2017) . In addition, the submission of only high quality candidate genes, continuous updating of a group’s submissions, removal of cases with ruled out/low priority candidate genes and removal or detailed annotation of solved cases (Table 2 ) will all promote efficient matchmaking using MME.
Several of the MME nodes have implemented additional features to help users track the large number of matches that currently require follow-up, such as those described for seqr (PRODUCTION: REFERENCE APPEARS IN THE SAME SPECIAL ISSUE (Pais et al., 2022) ), PhenomeCentral (PRODUCTION: REFERENCE APPEARS IN THE SAME SPECIAL ISSUE(Osmond, Hartley, Johnstone, et al., 2022) ), and GeneMatcher (PRODUCTION: REFERENCE APPEARS IN THE SAME SPECIAL ISSUE (Hamosh et al., 2022)) including auto-drafting emails as well as tracking match details, communications and the outcomes of matches within the database. An expanded use case for the MME is the secondary accumulation of additional cases after the initial publication of a novel disease gene relationship. Going forward, opportunities to support connections between investigators made across the MME beyond email notification will be considered, such as the equivalent of a gene ‘break-out’ room, to help these teams organize with less back-and-forth correspondence.