"variant" instead of "mutation" in genetics browser widget

[ElaineMcA]

@nikiforosk commented on Wed Aug 30 2017

Yesterday I met Jeff and we were looking at the Genetic associations for SNCA - Parkinson's: https://www.targetvalidation.org/evidence/ENSG00000145335/EFO_0002508?view=sec:genetic_associations

He mentioned that on the browser widget we should be using the term "variant(s)" instead of "mutation(s)":

The reason for this, as far as I understood, is that mutations are variants that have not been inherited by one's parent whereas what we show in these tracks are common variants from GWAS. Although many people use both terms interchangeably, this is not correct in Jeff's view.

We also noticed that on 2nd column of the Common diseases table we use the header "Variant" while the corresponding column on the Rare diseases is called "Mutation". Jeff was OK with that (although this seems to suggest that it's OK to use "Mutations" on the 2nd track of the widget - but not on the 1st track and the legend underneath). @iandunham any thoughts?

---

@iandunham commented on Wed Aug 30 2017

I think this is an oversight (we shouldn't have been using
  'mutations' in the browser for the common disease track). 

The distinction we are supposed to be using is "variant" for
  common base differences that exist in a population, and "mutation"
  for changes that have been shown to associated with a significant
  phenotypic change i.e. a rare disease or a cancer for somatic
  mutations.
There are shades of grey in this distinction though. So
  'mutation' is used in different ways across different communities
  (read the following if you want to become confused).
For the biochemist, the act of changing the chemical nature of a
  nucleotide is a mutation. So each change is caused by a mutation
  event, although later it might be called a variation.
For the population geneticist, the variants that are produced by
  mutation segregate in the population due to processes like
  selection and genetic drift.
For the classical geneticist a mutation has first of all a
  recognisable phenotype that you observe in your organism of
  interest, e.g. red/white eyes in flies, a rare disease in humans,
  and you then work back to identify the base change that causes the
  phenotype, which you refer to as the causal mutation. 

For the oncologist, you observe all the changes that occur in the
  somatic cells of a cancer and call them mutations even though some
  may not have phenotypes (e.g. driver and passenger mutations)

For the common disease geneticists, all differences in the base
  sequence between individuals in a population are variants, until
  such time as you can show that one is causal of a phenotypic
  change, and even then you might not call it a mutation....but a
  causal variant. 

So in summary the process of mutation gives rise to variants that
  might be called mutations if they have a strong phenotypic effect
  or are somatic (cancer). 

On 30/08/2017 15:45, nikiforosk wrote:

  Yesterday I met Jeff and we were looking at the Genetic
    associations for SNCA - Parkinson's:
    https://www.targetvalidation.org/evidence/ENSG00000145335/EFO_0002508?view=sec:genetic_associations
  He mentioned that on the browser widget we should be using the
    term "variant(s)" instead of "mutation(s)":

  The reason for this, as far as I understood, is that mutations
    are variants that have not been inherited by one's parent
    whereas what we show in these tracks are common variants from
    GWAS. Although many people use both terms interchangeably, this
    is not correct in Jeff's view.
  We also noticed that on 2nd column of the Common diseases table
    we use the header "Variant" while the corresponding column on
    the Rare diseases is called "Mutation". Jeff was OK with that
    (although this seems to suggest that it's OK to use "Mutations"
    on the 2nd track of the widget - but not on the 1st track and
    the legend underneath). @iandunham any
    thoughts?

  —
    You are receiving this because you were mentioned.
    Reply to this email directly, view it on GitHub, or mute the thread.

  {"api_version":"1.0","publisher":{"api_key":"05dde50f1d1a384dd78767c55493e4bb","name":"GitHub"},"entity":{"external_key":"github/opentargets/webapp","title":"opentargets/webapp","subtitle":"GitHub repository","main_image_url":"https://cloud.githubusercontent.com/assets/143418/17495839/a5054eac-5d88-11e6-95fc-7290892c7bb5.png","avatar_image_url":"https://cloud.githubusercontent.com/assets/143418/15842166/7c72db34-2c0b-11e6-9aed-b52498112777.png","action":{"name":"Open in GitHub","url":"https://github.com/opentargets/webapp"}},"updates":{"snippets":[{"icon":"DESCRIPTION","message":"\"variant\" instead of \"mutation\" in genetics browser widget (#188)"}],"action":{"name":"View Issue","url":"https://github.com/opentargets/webapp/issues/188"}}}

-- 

Cheers Ian

Ian Dunham M.A. D.Phil Scientific Director Open Targets European Bioinformatics Institute (EMBL-EBI) European Molecular Biology Laboratory Wellcome Genome Campus Hinxton Cambridge CB10 1SD United Kingdom

Tel: 01223 492636 FAX: 01223 494468 dunham@ebi.ac.uk

http://targetvalidation.org http://opentargets.org http://twitter.com @DunhamReal

---

@deniseOme commented on Thu Sep 07 2017

Jeff is not the first to be concerned with the way we interchangebly use 'variant' and 'mutation'. This was raised at the training UCL Alzheimers Res UK in July and discussed in one of our meetings where @iandunham asked me to list the places (yet to be done by me) where variants and mutations are used (the latter should be in rare disease track, and in cancer only). Ensembl calls mutations the variants coming from COSMIC (who mantains cancer gene census) and from The Human Gene Mutation Database (http://www.hgmd.cf.ac.uk/ac/index.php). Everything else is variants.

We can choose our definition based on @iandunham's comments above and stick with it. I will draft something and circulate. Once it's agreed we can have this as a FAQ (or tool tip in the Browser view and other places).

[LucaFumis]

Those labels (e.g. "Mutations in common diseases") are defined in https://github.com/opentargets/targetGenomeBrowser/blob/master/src/targetGenomeBrowser.js

Will then need @emepyc to update it in npm, or better to give some of us publish permission in npm.

[LucaFumis]

Also, just to summarise, are the changes as follow? In the browser tracks: 1) "Mutations in common diseases" : change to "Variants in common disease" 2) "Mutations in rare diseases" : keep as it is? In the browser legend: 3) "Mutation in NOD2 associated with Crohn's disease" : change to "Variant in NOD2 associated..." 4) "Mutation associated with Crohn's disease in other genes" : change to "Variant associated with..." In the tables: 5) keep "variant" in common disease table and "mutation" in rare diseases?

[deniseOme]

Yes, this is correct.

This is our FAQ: https://docs.targetvalidation.org/faq/variants-or-mutations

I will add more of @iandunham's explanation to the current FAQ.

[andrewhercules]

Just to follow-up from yesterday's front-end/UX meeting, it was decided that the changes will be made as summarised by @LucaFumis above:

"Mutations in common diseases" : change to "Variants in common disease" "Mutations in rare diseases" : keep as it is

In the browser legend:

"Mutation in NOD2 associated with Crohn's disease" : change to "Variant in NOD2 associated..." "Mutation associated with Crohn's disease in other genes" : change to "Variant associated with..."

In the tables:

Keep "variant" in common disease table and "mutation" in rare diseases table

[LucaFumis]

Deployed in version 3.7.4