Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Probable over-identification of BA.3 #584

Closed
shay671 opened this issue Apr 25, 2022 · 4 comments
Closed

Probable over-identification of BA.3 #584

shay671 opened this issue Apr 25, 2022 · 4 comments
Labels
pangoLEARN Issues related to pangoLEARN

Comments

@shay671
Copy link

shay671 commented Apr 25, 2022

A joint work with @talyash and @Boltyansky

At GISAID today we found 4,796 sequences identified as BA.3.

We extracted VCF from the sequences and examined for each of them the number of mutations corresponding to it in the consensus of each of the five variants in the Omicron lineage (BA.1-BA.5). The test was performed without deletion mutations.
For each of the sequences we examined which of the five variants had the maximal match :

<style> </style>
Variant #
BA.1 545
BA.2 3473
BA.3 715
BA.4 9
BA.5 2

So it seems that most of those might not be BA.3.

The analysis :
BA.3 from GISAID.xlsx

@hsnguyen
Copy link

I downloaded the GISAID complete BA.3 sequences and run the latest Pangolin on it.
The majority (85%) of these sequences were called Probable Omicron BA.3-like and they're mostly of low quality.

Count Mean #Ns
Probable 3522 6210
Default 594 70

So @aineniamh my feeling is that the rule for Probable BA.3 maybe a bit too relaxed thus overcalled.
Can you please double-check and consider making it more stringent or just remove the probable call, just like BA.4/5 for now.
Thanks,

@garfinjm
Copy link

We had a sequencing run with some major wetlab issues last week that I think shows this really starkly, these were all called Probable BA.3 by the latest StaPH-B Pangolin Docker image (4.0.6-pdata-1.6) using accurate analysis mode. I did some manual checking of sites that differ between BA.2 and BA.3 and these all seemed to match better (where there was sequence) with BA.2.

image

lineage conflict ambiguity_score scorpio_call scorpio_support scorpio_conflict scorpio_notes version pangolin_version scorpio_version constellation_version is_designated qc_status qc_notes note
BA.3 Probable Omicron (BA.3-like) 0.54 0.07 scorpio call: Alt alleles 29; Ref alleles 4; Amb alleles 21; Oth alleles 0 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.42 scorpio called lineage BA.3
BA.3 Probable Omicron (BA.3-like) 0.46 0.04 scorpio call: Alt alleles 25; Ref alleles 2; Amb alleles 26; Oth alleles 1 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.37 scorpio called lineage BA.3
BA.3 Probable Omicron (BA.3-like) 0.56 0.09 scorpio call: Alt alleles 30; Ref alleles 5; Amb alleles 19; Oth alleles 0 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.38 scorpio called lineage BA.3
BA.3 Probable Omicron (BA.3-like) 0.46 0.02 scorpio call: Alt alleles 25; Ref alleles 1; Amb alleles 28; Oth alleles 0 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.44 scorpio called lineage BA.3
BA.3 Probable Omicron (BA.3-like) 0.46 0.02 scorpio call: Alt alleles 25; Ref alleles 1; Amb alleles 28; Oth alleles 0 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.45 scorpio called lineage BA.3
BA.3 Probable Omicron (BA.3-like) 0.39 0.04 scorpio call: Alt alleles 21; Ref alleles 2; Amb alleles 31; Oth alleles 0 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.63 scorpio called lineage BA.3
BA.3 Probable Omicron (BA.3-like) 0.54 0.07 scorpio call: Alt alleles 29; Ref alleles 4; Amb alleles 21; Oth alleles 0 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.43 scorpio called lineage BA.3
BA.3 Probable Omicron (BA.3-like) 0.54 0.07 scorpio call: Alt alleles 29; Ref alleles 4; Amb alleles 21; Oth alleles 0 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.43 scorpio called lineage BA.3
BA.3 Probable Omicron (BA.3-like) 0.54 0.07 scorpio call: Alt alleles 29; Ref alleles 4; Amb alleles 21; Oth alleles 0 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.42 scorpio called lineage BA.3
BA.3 Probable Omicron (BA.3-like) 0.54 0.06 scorpio call: Alt alleles 29; Ref alleles 3; Amb alleles 21; Oth alleles 1 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.42 scorpio called lineage BA.3
BA.3 Probable Omicron (BA.3-like) 0.54 0.07 scorpio call: Alt alleles 29; Ref alleles 4; Amb alleles 21; Oth alleles 0 SCORPIO_v0.1.8 4.0.6 0.3.17 v0.1.8 FALSE fail Ambiguous_content:0.43 scorpio called lineage BA.3

We have yet to see any BA.3 in MN, so seeing this many on one run (that aren't geographically clustered at the county level or epi-linked) threw up some red flags.

@corneliusroemer corneliusroemer added the pangoLEARN Issues related to pangoLEARN label Apr 28, 2022
@nknox
Copy link

nknox commented Apr 30, 2022

We are seeing the same thing - we only have one true BA.3 positive

image

Additive parent+child calls might not be stringent enough?
Default BA.3 call: min alt 35
Probable BA.3 call: min alt 18

image

@aineniamh
Copy link
Member

Thanks for flagging- I've updated the constellation definition to remove the probable call now in version 0.1.9 (https://github.com/cov-lineages/constellations/releases/tag/v0.1.9). This should resolve your issue!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
pangoLEARN Issues related to pangoLEARN
Projects
None yet
Development

No branches or pull requests

6 participants