GAW 17 data comes late this year. So people were rushing to make the coming deadline. We proposed a rather conservation method to analyze the data, however, the results do not look good in the first place. It took a while to observe from every aspect that it doesn't turn out the signal, with the accompany of the feeling of missing the deadline.
As more and more tests turned out not to be positive. I started to examine and analyze all the possible reasons. Though, all types of analysis look reasonable, but people don't have good ways to solve these problems. For example the influence of low heritability to the mapping of snp variations especially including the rare ones in NGS to traits.
This morning, I felt I may need to start a series of tests that may enlarge the signal units. Instead of genes, maybe signal pathways, since so many genes have only one or two rare variants, and it doesn't make sense to get anything from them.
Then, may change the view of extracting information. Not only use one rule, but may use multiple rules to collect information. Cross my figures...