griffithlab / GenVisR

Genome data visualizations
Creative Commons Zero v1.0 Universal
209 stars 62 forks source link

Assign translational effect to custom variant_class values #296

Closed u9090 closed 7 years ago

u9090 commented 7 years ago

Hi,

It seems that only 'Silent' mutations are assigned a 'Synonymous' translational effect on the mutation burden plot.

Is there a way to assign translational effects on a custom list of variant class?

For example to also assign 'Intron', 'UTR' and 'Flanking regions' a synonymous effect.

zlskidmore commented 7 years ago

Unfortunately not at this time, I do plan to add the option of coloring bars in the subplots by mutation type but that functionality is not ready yet. leaving this open as an enhancement

u9090 commented 7 years ago

Sorry, but it seems to me that counting a mutation occurring in an intron, UTR, flanking or intergenic region (i.e. non-coding regions) as non-synonymous (i.e. ending up in a protein sequence change) is wrong. It is not an enhancement.

zlskidmore commented 7 years ago

@lavv2 sorry, I was only thinking in terms of linking the existing issue which is currently classified as an enhancement, however, I of course agree with you, adding bug label instead. Do you have any thoughts on the proper way to go about this? There is no way to know the effect an intron will have without additional information, for example an intronic mutation could cause a some type of splice effect altering the protein in which case it should be non-synonymous or it could not and as you say should be classified as synonymous.

I think the way to go about this is to remove the synonymous/nonsynonymous classification and add a parameter to either just do a regular barplot, or alternatively a stacked barplot based on the variant_class. Do you agree?

u9090 commented 7 years ago

Thanks for considering this issue.

Not sure a stacked barplot based on the 'variant_class' would add more useful information than what is already in the main plot. The added value of the Translational Effect info is the concise and larger clustering of variant types I guess.

The issue may come from the format of the custom file type input. I personally use Annovar which defines slightly different variant classes compared to MAF or MGI format (see the 2 tables here: http://annovar.openbioinformatics.org/en/latest/user-guide/gene/#output-file-1-refseq-gene-annotation). For splicing, if the mutation is within 2-bp (default) to a splice site, it is assumed that it will have a non-synonymous effect as you mentioned. But this region length could be different for other user purposes.

From your side I am not sure you can do much to predict user annotation format. To my perspective (and not to make yourself crazy) the best way would be to let the user deal with what is synonymous and not.

zlskidmore commented 7 years ago

that makes sense, I am adding support for additional filetypes, I will add annovar to the list. But I will add in the functionality to let a user decide what is synonymous while i'm at it if filetype is custom

zlskidmore commented 7 years ago

Work for this is being done here https://github.com/griffithlab/GenVisR/tree/S4_conversion

zlskidmore commented 7 years ago

fixed in 72d045fcb86c57b08180ac1fbd0c2b0833e0cd01

snashraf commented 6 years ago

I have used VCF2MAF for converting my VCF to MAF file and then I am trying to use GenVisR for plotting and its WGS data from Non-Cancer sample but wanted to plot waterfall plots for few ganes. Is their a way to drop all non supporting annotation from genVisR ? Detected an invalid mutation type, valid values for maf are: Nonsense_Mutation, Frame_Shift_Ins, Frame_Shift_Del, Translation_Start_Site, Splice_Site, Nonstop_Mutation, In_Frame_Ins, In_Frame_Del, Missense_Mutation, 5'Flank, 3'Flank, 5'UTR, 3'UTR, RNA, Intron, IGR, Silent, Targeted_Region, NA

zlskidmore commented 6 years ago

I would recommend setting fileType="Custom" and using variant_class_order to set valid annotations. It will automatically fail if all the annotation types are not set but otherwise it should be fine.

Let me know if that works for you

snashraf commented 6 years ago

Is their a way to ignore all which is not mentioned in variant_class_order rather than just failing it ?

On Thu, Jan 25, 2018 at 6:48 PM, Zachary Skidmore notifications@github.com wrote:

I would recommend setting fileType="Custom" and using variant_class_order to set valid annotations. It will automatically fail if all the annotation types are not set but otherwise it should be fine.

Let me know if that works for you

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/griffithlab/GenVisR/issues/296#issuecomment-360507342, or mute the thread https://github.com/notifications/unsubscribe-auth/AEJUN-49rmlsNmckqxC9Xfxq5Ny12JaHks5tOKJrgaJpZM4MaDHy .

-- Syed Najeeb Ashraf

snashraf commented 6 years ago

OK Thanks !! I did it by filtering the datasets only !! Thanks

On Thu, Jan 25, 2018 at 9:18 PM, Syed Najeeb Ashraf < syednajeebashraf@gmail.com> wrote:

Is their a way a ignore all which is not mentioned in variant_class_order rather tha just failing it ?

On Thu, Jan 25, 2018 at 6:48 PM, Zachary Skidmore < notifications@github.com> wrote:

I would recommend setting fileType="Custom" and using variant_class_order to set valid annotations. It will automatically fail if all the annotation types are not set but otherwise it should be fine.

Let me know if that works for you

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/griffithlab/GenVisR/issues/296#issuecomment-360507342, or mute the thread https://github.com/notifications/unsubscribe-auth/AEJUN-49rmlsNmckqxC9Xfxq5Ny12JaHks5tOKJrgaJpZM4MaDHy .

-- Syed Najeeb Ashraf

-- Syed Najeeb Ashraf