Add P/LP or Conflicting option to the Pathogenicity section

[lynnpais]

Is your feature request related to a problem? Please describe.
Variants that have a pathogenic or likely pathogenic classification AND a VUS classification in ClinVar are classified as Conflicting, and do not meet the current override criteria in our standardized searches.

Describe the solution you'd like
Similar to how we have the VUS or Conflicting, add another option - P/LP or VUS Conflicting.

Describe alternatives you've considered
NA

Additional context
Discussed with the seqr and analyst teams.

[lynnpais]

Add to the override and saved searches

[hanars]

@lynnpais just confirming that if there is a "Uncertain Significance" classification that should be returned even if theres no path or likely path (e.g. https://seqr.broadinstitute.org/variant_search/variant/X-48957513-G-T/family/F057673_rgp_3579)

[lynnpais]

I should've mentioned that we'd like a new category of P/LP or VUS, so that it can added to our standard restrictive searches.

For example, the override for restrictive searches includes P or LP variants only. Hence variants like this PPP2R5D (marked conflicting because it has P, LP and a VUS classifications), are not being returned in a restrictive dominant search - https://seqr.broadinstitute.org/variant_search/results/57b39e964fbf9b1048bd2f983ab14e25?page=1&sort=in_omim

[hanars]

So the current checkbox options for clinvar are

• Pathogenic (P)
• Likely Pathogenic (LP)
• VUS or Conflicting
• Likely Benign
• Benign (B)

These existing options are sufficient to generally search for "P/LP or VUS" by just choosing the top 3 so I do not think a new category for that is necessary or appropriate.

What this ticket is proposing, as was discussed, is adding a new category to specifically search against the conflicting pathogenicities themselves - so it would match only on variants with conflicting pathogenicities and only if at least of of the conflicting assertions is "P/LP or VUS". This would be added as a new option to the list of options, and we would add it to our standard searches, and when this is included in search it would override the frequency/quality/annotations the same as P/LP currently do

What I was trying to clarify though is whether you actually want to include VUS in this search or whether you actually onyl want this category to include P/LP - to me it seems like something like the example I sent is not something the analysts would really want to have overriding frequency and quality filters or bloating restrictive searches. The example you sent would be included in this new search category regardless because it has P and LP pathogenicities - my question was really just scoped to how we want to handle the VUS part of "P/LP or VUS"

[lynnpais]

Got it, thanks for clarifying. And yes, we wouldn't want VUS returned if theres no path or likely path such as in the example you shared.

[hanars]

this is now live