To expand on this, the team is launching a new A/B testing framework today that makes it easier to run variant tests and try different push notification messages.
Finally, we performed two hypothesis driven related conditional tests (independent effect and sole variant tests) to try to identify which variant, or variants, is solely and independently associated with the disease [45].
Among them, 8 were overlapped by AraNet either through common variant tests (AP1, FCA, FRI, and FLC) or through rare variant tests (FLM, FY, BAS1, and SPL5).
Six other genes (two through common variant tests and 4 through rare variant tests) were determined to be interesting for follow-up studies because they all had top 30 significant tests and supporting evidence from function prediction.
The evaluation of different meta-analysis approaches of rare variant tests is an active field of study.
Just as with single variant tests, we recommend generating QQ plots to summarize association results across the genome and ensure test statistics are well behaved.
Since the rare allele of individual rare variants are observed so infrequently, single variant tests of association will be underpowered for all but the most highly penetrant alleles.
The analysis of rare variants across the genome requires a more stringent significance threshold that takes into account single-point common and rare variant tests as well as burden tests.
Single variant tests were performed under an additive model, and the empirical variance (the −e option) was used to ensure its validity as a test of association in the pedigree.
In the current study, one multiple common variant test and three pooled rare variant tests were examined to determine the significance of the gene fragments and compare the performance of these tests.
For these rare variants, single variant tests can lack power, and association tests that group rare variants by gene or functional unit are favored (Li and Leal, 2008; Lin and Tang, 2011; Madsen and Browning, 2009; Price et al., 2010; Wu et al., 2011).
