Thread by @NathanGrubaugh

UPDATE: GISAID put a fix in place to accept these sequences, and any sequence with a known FRAMESHIFT, w/o confirmation. 🍻👏

Here's what you need to do. Its pretty easy. 🧵 (1/6)

If you go to the Upload > Batch Upload tab, you should see the new "Confirmation options" dropdown menu highlighted in yellow. (2/6)

Now you have options!

The Default is as before, any sequencing with FRAMESHIFTS, like B.1.621, will be kicked back for confirmation and resubmission.

I don't recommend using the default. (3/6)

The third option should be left for the pros.

It means that GISAID will accept all of your FRAMESHIFTS w/o curation. As Illumina is known for indels in homopolymer regions, errors do occur.

If you are not verifying these in your sequences, please do not use this option. (4/6)

The "notify me only about NOT PREVIOUSLY REPORTED FRAMESHIFTS..." option in the middle is the real fix here.

This means that GISAID will not make you confirm/resubmit B.1.621 sequences that contain the real 4nt deletion in ORF3a.

USE THIS! (5/6)

👆 This also means that other real deletions that are previously reported will also be immediately uploaded, alleviating an important data bias in our system.

**THANK YOU** GISAID for hearing our concerns and creating a real solution to this problem. 👏🍻 (6/6)

was in response to this. For some reason the "quote tweet" didn't work.

https://twitter.com/NathanGrubaugh/status/1420797454782402560