impact of mutations in splicing sites

[david.tamborero]

thanks for the answer!

I think that I am going to score as 'highly deleterious' those mutations in 2 last/ 2 initial bases of introns in canonical splice sites, as I am currently doing with mutations in exons introducing a stop codon or a frameshift indel. For those mutations in other splice regions of the intron, for the moment I will consider them as having unknown effect, since I am not able to find a reliable score to classify their deleterious potential, and actually they should be not well catalogued -I guess- since I am dealing with exome seq data.

Now I will move towards the assessment of how mutations in exons can create cryptic splice junctions. Absolutely, I will check skippy, it looks very nice!

thanks again,
best
David

[m_two]

If you are interested in de novo splice site potential you might be interested in this:
SKIPPY_-_Exonic_Splice_Modulator_Scoring

Web query:

http://research.nhgri.nih.gov/skippy/input.shtml

Home:

http://research.nhgri.nih.gov/skippy/index.shtml

Pub:
Woolfe, A., Mullikin, J., and Elnitski, L. 2009. Genomic features defining exonic variants that modulate splicing. Genome Biology. 11:R20 [PubMed]

http://genomebiology.com/content/11/2/R20

See: Table S1 – List of 87 Synonymous and Missense variants that cause exon skipping ,used for analysis in this study. The variants are located in 40 genes and 47 individual ,exons. Variant locations are from human genome assembly hg18. References can be ,found in the main text of the paper.

Additional file 1. List of 87 synonymous and missense splice-affecting genome variants (SAVs) that cause exon skipping used for analysis in this study. The variants are derived from [12,13,37,41,44,47,48,54-56,65-103].

Format: PDF Size: 117KB Download file

http://genomebiology.com/content/supplementary/gb-2010-11-2-r20-s1.pdf

[m_two]

Ensembl VEP annotates mutations affecting the 5' and 3' and last two bases of the annotated introns as splice_donor_variant / splice_acceptor_variant, it does not predict de novo splice sites.

To further filter the annotation you can try these four things:

Eliminate variants exclusive to transcripts that are known pseudogenes or undergo NMD.

Grab the nucleotide conservation scores from UCSC, real splice sites should have very high conservation scores.

Eliminate splice variants that affect noncanonical GT AG splice sites.

Filter common dbSNPs identified in 1000 Genomes data and NHLBI GO Exomes

[david.tamborero]

hi,

I would like to score the impact of mutations occurring in gene regions that the variant effect predictor state as 'splice_variant'. Note that at this moment I do not want to predict whether the mutations create a cryptic splice site.

I guess that there are some consensus sequences for both donor and acceptors, but I assume that it is more complicated that to construct the weight matrix by myself and check how it fits the wild type vs the mutant.

Actually, I'm a bit lost. I found the Human Splice Finder and the MaxEntScan, but I am not sure of how to interpret their results, and -mainly- I would need some tool that can be run by script, since them are only available via manual entering of the data in the web.

many thanks,
david