Extend the query to support region queries

[juhtornr]

Can you @jrambla add description based on your presentation to ELIXIR Beacon strategic group?

[mbaudis]

@juhtornr @jrambla I've added some notes & an example.

Region queries are used to determine the existence of any/all variants in a genomic range. A typical example would be the determination of variants n the CDR of a gene of interest. In this example, all variants with single nucleotide alternateBases in the CDR of the EIF4A1 gene in the DIPG childhood brain tumor dataset are being retrieved:

https://beacon.progenetix.org/beaconplus-server/beaconresponse.cgi?datasetIds=dipg&referenceName=17&assemblyId=GRCh38&startMin=7572826&endMax=7579005&referenceBases=*&alternateBases=N

Special parameters:

• datasetIds=dipg
  • limit to the DIPG dataset
• startMin=7572826
• endMax=7579005
  • In this proposed form, the startMin and endMax parameters are used to indicate the extent of the queried region. This is in contrast to using start and end, which should be considered parameters for precise positions (e.g. a variant from start to end, not in the corrsponding range). However, this will need agreement & documentation.
• referenceBases=*
  • any reference base (wildcard query)
  • also specific replacements could be queried for
• alternateBases=N
  • The current API does not allow wildcards, but just undefined bases of specific count. Here, the query would match e.g. A>G, but not A>GA (one would have to use NN for this).

TODO

• agree on position parameters used for ranges
• think about wildcard options for alternateBases (in the sense of "*" or "/N+?/").
• decide about a variant disambiguation strategy for the response
  • Variants in handover & post-processing?
  • Variants in response with variant specific counts?

[jrambla]

Hi, I'll leverage this use case to test the new focused queries. Thus, we'll remove any non required parameter in the query. Back to you soon (well, after holidays) Written from a mobile device, disregard brevity and typos, please. Escrit des d'un mòbil, disculpeu la brevetat i les errades. Escrito desde un móvil, disculpad la brevedad y las erratas.

…

________________________________ From: Michael Baudis <[email protected]> Sent: Tuesday, December 11, 2018 1:02:55 PM To: ga4gh-beacon/specification Cc: jrambla; Mention Subject: Re: [ga4gh-beacon/specification] Extend the query to support region queries (#235) @juhtornr<https://github.com/juhtornr> @jrambla<https://github.com/jrambla> I've added some notes & an example. Region queries are used to determine the existence of any/all variants in a genomic range. A typical example would be the determination of variants n the CDR of a gene of interest. In this example, all variants with single nucleotide alternateBases in the CDR of the EIF4A1 gene in the DIPG childhood brain tumor dataset are being retrieved: https://beacon.progenetix.org/beaconplus-server/beaconresponse.cgi?datasetIds=dipg&referenceName=17&assemblyId=GRCh38&startMin=7572826&endMax=7579005&referenceBases=*&alternateBases=N Special parameters: * datasetIds=dipg * limit to the DIPG dataset * startMin=7572826 * endMax=7579005 * In this proposed form, the startMin and endMax parameters are used to indicate the extent of the queried region. This is in contrast to using start and end, which should be considered parameters for precise positions (e.g. a variant from start to end, not in the corrsponding range). However, this will need agreement & documentation. * referenceBases=* * any reference base (wildcard query) * also specific replacements could be queried for * alternateBases=N * The current API does not allow wildcards, but just undefined bases of specific count. Here, the query would match e.g. A>G, but not A>GA (one would have to use NN for this). TODO * agree on position parameters used for ranges * think about wildcard options for alternateBases (in the sense of "*" or "/N+?/"). — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#235 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AHsiOkvN548lOe4Xbaam2PB3gKFuFRM9ks5u357vgaJpZM4YpWzc>.

[mbaudis]

@jrambla @juhtornr Great - I'll be happy o help, developing this further. For now, the query example still works ...
@teemukataja This also reports the short-form "digest" of the matched variants in the variantResponses (e.g. "DIPG_V_MAF_17_7578406_C_T"), since we've set the limit to "20" distinct variants for full reports. If you limit the target region you'll get a full report from the variants (slowish...): https://beacon.progenetix.org/beaconplus-server/beaconresponse.cgi?datasetIds=dipg&referenceName=17&assemblyId=GRCh38&startMin=7576826&endMax=7577200&referenceBases=*&alternateBases=N