More Autism Genetic Variants

In my previous post Some of My Autism Related Genetic Variants, I gave a list of potential autism related variants that I found in my genome sequencing results. In that analysis I focused on coding variants. These are variants that directly change the sequence of nucleotides that encode the protein. However there are many other variants both in and out of genes that can have a deleterious impact on protein function and thus cause disorders.

To delve deeper in to these variants I made use of an algorithm called Linsight. Linsight provides a score that indicates how likely non coding variants are to have a functional impact. If you want to understand how it does this then you should read the paper Fast, scalable prediction of deleterious noncoding variants from functional and population genomic data, but it is enough to know that higher scores indicate a higher likelihood of functional impact.

I filtered my data by Sfari’s gene list (download here), Gnomad frequency < 0.0001 and Linsight > 0.9 and I got the following 8 variants. That seemed a reasonable number to do some further research on.

Gene	Sfari Category	CADD 1.7	Linsight	Inherited From	Variant Type	GnomAD4	Varsome
TBR1	1	20.1	0.989705	Dad	2kb upstream	chr2,161416205,C,G	chr2161416205,C,G
DPYSL2	1	32	0.971332	Mum	missense	chr8,26514492,C,G	chr8,26514492,C,G
TCF4	1	17.67	0.968867	Dad	intron	chr18,55355434,T,G	chr18,55355434,T,G
CNTNAP2	2	18.61	0.973429	Mum	intron	chr7,146912021,C,T	chr7,146912021,C,T
TBC1D5	2	22.1	0.971925	Dad	intron	chr3,17352884,C,A	chr3,17352884,C,A
CCDC91	2	17.31	0.955174	Dad	intron	chr12,28136413,G,T	chr12,28136413,G,T
NFIB	2	15.36	0.910331	Dad	intron	chr9,14286337,A,G	chr9,14286337,A,G
MSRA	3	13.08	0.958318	Mum	intron	chr8,10191391,T,G	chr8,10191391,T,G

I was intrigued to find that one of the variants that appeared in this list was a coding variant I found in my previous analysis, a DPYSL2 missense variant. Looking into the DPYSL2 gene I discovered that this gene has several transcripts of varying length. All of the pathogenic variants that have been identified thus far are in the part of the DPLYSL2 gene that falls in all transcripts. My variant however is in exon 1 of the longest transcript and so it is only in that single transcript. Most likely that would mean that any variants in this exon would have a milder impact, if any, than the pathogenic variants found to date. While this is interesting it does not explain the rather high Linsight score.

Further research showed that my variant also falls in the upstream region of nearby gene PNMA2. I found another tool online, Regulation Spotter, that provides a score indicating how likely a variant is to have a functional impact on a regulatory region. (Variants have to use hg19 coordinates for this tool.) In addition to the score itself, information on how that score is reached is provided. For my variant Regulation Spotter shows that it is in the promotor region for PNMA2 and so may affect the functioning of that protein. PNMA2 is not associated with Autism or any other neurodevelopmental disorder currently so while I do think this variant is interesting, at present there is minimal evidence that it may play a role in Autism.

One other variant particularly stood out while I was doing further research on the remaining 7 variants, the TBR1 variant. TBR1 is highly associated with Autism, it was in fact one of the earliest genes to be associated with Autism. My variant falls in the upstream region of the gene. Regulation Spotter indicates that this variant falls within the promoter region for the TBR1 gene and so could effect the protein. At present no variants in the promoter region of TBR1 have been classified at all, but I do think this will be something to keep an eye on going forward.