Tracing Mutations Back to Lineage
The Python Outbreak API can be queried in order to determine which lineages a mutation has been found in. After collecting a sample and determining what sequences are present, we may have a list of several SARS-CoV-2 mutations that we can immediately say are characteristic of a specific variant. However in some cases, we also may have a mutation that is relatively uncommon in most other samples. For example, we can look at small data sample consisting of 10 mutations: (S:A67V, S:DEL69/70, S:E484A, S:N501Y, S:T572N, S:D614G, S:G142D N:S2Y, S:Q52R, E:L21F, S:G593D). We’ll want a way to find more details about any mutation collected, such as whether the mutation has been collected before, when, and where that mutation came from.
To start, the mutations_by_lineage() function allows us to look at the clinical prevalence of a mutation and see which lineage it most likely belongs to. Let’s try it for E:L21F:
# Perform authentication if you haven't already
from outbreak_data import authenticate_user
authenticate_user.authenticate_new_user()
# Import outbreak_data package
from outbreak_data import outbreak_data as od
lin1 = od.mutations_by_lineage(mutation='E:L21F')
print(lin1)
pangolin_lineage lineage_count mutation_count proportion \
0 ba.2 1228296 560 0.000456
1 b.1.1.7 1155169 844 0.000731
2 ba.1.1 1046121 268 0.000256
3 ay.4 861521 526 0.000611
4 ba.1 439838 49 0.000111
... ... ... ... ...
400 ba.2.77 63 48 0.761905
401 ba.5.2.54 55 2 0.036364
402 b.1.616 39 3 0.076923
403 b.1.1.386 20 1 0.050000
404 b.1.1.400 20 20 1.000000
proportion_ci_lower proportion_ci_upper
0 0.000419 0.000495
1 0.000683 0.000781
2 0.000227 0.000288
3 0.000560 0.000664
4 0.000083 0.000146
... ... ...
400 0.646596 0.853783
401 0.007632 0.111568
402 0.022142 0.191265
403 0.005449 0.210819
404 0.883361 0.999976
[405 rows x 6 columns]
This mutation has clearly been seen before in some previous lineages. We might be able recognize that most of the mutations in our list have been detected in older variants, as well as Omicron. However, S:G593D is relatively uncommon in most other samples. We can easily find out where and when it was last detected:
>>> lin2 = od.mutations_by_lineage(mutation='S:G593D')
>>> print(lin2)
pangolin_lineage lineage_count mutation_count proportion \
0 xbb.1 28205 1 0.000035
proportion_ci_lower proportion_ci_upper
0 0.000004 0.000166
>>> last_seen = od.collection_date('xbb.1', 'S:G593D')
>>> print(last_seen)
Values
date 2022-12-12
date_count 1
According to our data, S:G593D has only been detected once in a single sequence belonging to the xbb.1 lineage. The last time it was collected was back on December 12, 2022.
Additionally mutations_by_lineage allows us to find out if there is a lineage where several mutations overlap. Selecting 7 of the mutations from our original list yields one lineage with all of these mutation characteristics:
>>> lin3 = od.mutations_by_lineage(mutation='S:A67V, S:DEL69/70, S:E484A, S:N501Y, S:T572N, S:D614G, S:G142D')
>>> print(lin3)
pangolin_lineage lineage_count mutation_count proportion \
0 ba.1.19 4587 1 0.000218
proportion_ci_lower proportion_ci_upper
0 0.000024 0.001019
Here we see that the only lineage that contains all 7 mutations is ba.1.19.