Hi everyone,
I am trying to find a way to understand the effect of a splice site mutations (GT--AG or AC--CT) in mouse genome. I want to know if there is a manual method or tool(s) which can predict the effect of splice site mutations in MOUSE genome.
Most of the tool I saw are designed for Human. It would be helpful if anyone can suggest similar tool(s) or method for Mouse.
Is there any tools like MutationTaster or any other for mouse?
I tried to use Spliceman, but I am not sure that one is the correct tool for my query. Also I didnt get any exact effect of how that splice mutation will be affecting transcription.
Also is there a possibility to see the effect of these mutations on my RNA-seq data of mouse.
I will appreciate any help.
Thanks
Ankit
The original version of Spliceman was computed across 11 species and the tool has since been improved, though the latest version only has hg19 and hg38. Spliceman isn't exactly interested in splice site mutations, as those are typically predicted to alter splicing. Most people are interested in mutations outside of the canonical signals and how they may have an effect in splice site recognition.
I don't think splicing tools are meant to predict the effect of mutations in transcription.
Hi Eric,
Thanks for the reply.
I want to know if a splice variant in mouse has deleterious effects. I am looking for the effect in my RNA-seq data (normalised read counts), but not able to find anything. So I thought may be a database for mouse splice variants or tools can help me to get a bit of idea. But I couldn't find such tools for mouse.
Thanks
If the observed mutation changes a canonical splice site to non-canonical, splice site recognition is often altered. If such a change does exist, you should observe it in your RNA-Seq data, either via the expression of splice junction or isoform. If the effect is big enough and it triggers NMD, you should also see a change in gene expression.
One of the follow-up papers of Spliceman (https://genome.cshlp.org/content/early/2015/12/11/gr.181008.114.full.pdf; Figure 1C) looked at these mutations across 20+ species and found that most signals, especially canonical, are conserved. Splicing tools in mouse is certainly lacking in the field, but given splicing is generally conserved across species, many of the tools developed for human should be applicable to other species as well, especially if you're interested in exonic variants.
If you're interested in exonic splicing enhancers, there is RESCUE-ESE (http://genes.mit.edu/burgelab/rescue-ese/), which was pre-computed in mouse. The idea is: if your mutation happens to disrupt a functional ESE, it increases the likelihood that splicing is affected. Functional is the keyword here.
If your region is conserved between human and mouse, tools like http://www.umd.be/HSF/ might be useful.
There are many ways to do this, you might just need to be creative. But at the end of the day, as splicing is so complex, experimental validations in conjunction with computational analyses are required.
If you have a set of canonical splice variants of each gene, then you could separate out only the non-canonical variants and look for the mutations. Once you find the mutations, you could check if the variant and its effects have been reported.
Hi Vinay,
Thanks for the reply.
I extracted only the canonical transcripts and marked splice sites in them. I also know where the mutations (SNPs and Indels) are present in those splice sites . But how should I check the effect of these splice variants. The number of variants are high. So it would be nice if I get an idea of databases, tools or any other method to get idea of probable effect using a batch query of MOUSE splice variants.
In my answer here, I provide some links to some splice-site prediction tools: A: pathogenicity predictors of cancer mutations
Ankit,
At what level do you want to understand the effect of the mutations? If it's at the protein level, you could simply translate your sequences and perform a BLAST. In cases where you get a high query cover and identity would carry mutations in regions other than the splice sites. In case of splice variants (intron inclusion or exon exclusion), the query cover and the identity will also be lower. But if the non-canonical variants show 100% query cover and identity, it would mean that these variants have been reported and studied.
Hope this is helpful.
Hi Vinay, I didn't get it. Should I translate the sequence containing mutation?
Ankit,
If you want to study the effects of the mutation at the protein level, you could translate and perform a BLAST. If sequences aren't reported, those would be novel variants.