Thread by @jbloom_lab

Below is brief analysis of HA mutations in two recent cases of H5N1 influenza in humans w contact w dairy cattle in California.

Summary is that while virus continues to evolve, nothing about HA mutations in these human cases is obviously alarming.

As background, CDC reported several recent cases of H5 influenza in California.

CDC and California DOH recently shared sequences of two of these cases via GISAID.
cdc.gov/media/releases…

California human cases share two HA mutations relative to "consensus" dairy cattle virus HA:

D95G & S336N in H3 numbering (D88G & S320N in H5 numbering; D014G & S336N in sequential numbering).

Both these mutations also in some dairy cattle HAs, so not unique to human cases.

One of the California human cases also has A218S (a la A214S or A230S). This mutation appears to be unique to this case.

Based on our deep mutational scanning, none of these mutations have an obvious strong effect on HA antigenicity or receptor usage ()

(As always, caveat is that there could be some effect we are missing.)dms-vep.org/Flu_H5_America…

Here are the locations of the mutations on the HA structure. Two are on the head, one is in the stalk.

The California human cases *lack* the A160T (a la A156T or A172T) mutation that was present in a recent human case from Missouri and is known to cause a substantial antigenic change.

Note above thread only discusses HA mutations, not mutations in other parts of genome which could be relevant.

Also, HA mutations could have some important effect I am missing (see for more discussion)

But I see nothing in HA mutations of obvious alarm

Here is link to interactive HA structure where you can look at mutations and associated deep mutational scanning data: dms-viz.github.io/v0/?data=https…

I apologize for typo here, it is S323N in H3 numbering, not S336N. The table attached to the above post is correct.