Skip to content

Sequence lineage information extracted from RKI sequence data repo

License

Notifications You must be signed in to change notification settings

corneliusroemer/desh-data

Repository files navigation

Pango lineage information for German SARS-CoV-2 sequences

This repository contains a join of the metadata and pango lineage tables of all German SARS-CoV-2 sequences published by the Robert-Koch-Institut on Github.

The data here is updated every hour, automatically through a Github action, so whenever new data appears in the RKI repo, you will see it here within at most an hour.

The resulting dataset can be downloaded here, beware it's currently around 50MB in size: https://raw.githubusercontent.com/corneliusroemer/desh-data/main/data/meta_lineages.csv

Omicron share plot

Type N means representative surveillance. Type X means unknown, but since this is unlikely to be heavily targeted and includes quite a number of labs I include it now in the main plot (hence type NX).

Omicron Logit Plot

Omicron Logit Plot

Omicron share by zip code area

Description of data

Column description:

  • IMS_ID: Unique identifier of the sequence
  • DATE_DRAW: Date the sample was taken from the patient
  • SEQ_REASON: Reason for sequencing, one of:
    • X: Unknown
    • N: Random sampling
    • Y: Targeted sequencing (exact reason unknown)
    • A[<reason>]: Targeted sequencing because variant PCR indicated VOC
  • PROCESSING_DATE: Date the sample was processed by the RKI and added to Github repo
  • SENDING_LAB_PC: Postcode (PLZ) of lab that did the initial PCR
  • SEQUENCING_LAB_PC: Postcode (PLZ) of lab that did the sequencing
  • lineage: Pango lineage as reported by pangolin
  • scorpio_call: Alternative, rough, variant as determined by scorpio (part of pangolin), this is less precise but a bit more robust than pangolin.

Excerpt

Here are the first 10 lines of the dataset.

IMS_ID,DATE_DRAW,SEQ_REASON,PROCESSING_DATE,SENDING_LAB_PC,SEQUENCING_LAB_PC,lineage,scorpio_call
IMS-10294-CVDP-00001,2021-01-14,X,2021-01-25,40225,40225,B.1.1.297,
IMS-10025-CVDP-00001,2021-01-17,N,2021-01-26,10409,10409,B.1.389,
IMS-10025-CVDP-00002,2021-01-17,N,2021-01-26,10409,10409,B.1.258,
IMS-10025-CVDP-00003,2021-01-17,N,2021-01-26,10409,10409,B.1.177.86,
IMS-10025-CVDP-00004,2021-01-17,N,2021-01-26,10409,10409,B.1.389,
IMS-10025-CVDP-00005,2021-01-18,N,2021-01-26,10409,10409,B.1.160,
IMS-10025-CVDP-00006,2021-01-17,N,2021-01-26,10409,10409,B.1.1.297,
IMS-10025-CVDP-00007,2021-01-18,N,2021-01-26,10409,10409,B.1.177.81,
IMS-10025-CVDP-00008,2021-01-18,N,2021-01-26,10409,10409,B.1.177,
IMS-10025-CVDP-00009,2021-01-18,N,2021-01-26,10409,10409,B.1.1.7,Alpha (B.1.1.7-like)
IMS-10025-CVDP-00010,2021-01-17,N,2021-01-26,10409,10409,B.1.1.7,Alpha (B.1.1.7-like)
IMS-10025-CVDP-00011,2021-01-17,N,2021-01-26,10409,10409,B.1.389,

Suggested import into pandas

You can import the data into pandas as follows:

#%%
import pandas as pd

#%%
df = pd.read_csv(
    'https://raw.githubusercontent.com/corneliusroemer/desh-data/main/data/meta_lineages.csv',
    index_col=0,
    parse_dates=[1,3],
    infer_datetime_format=True,
    cache_dates=True,
    dtype = {'SEQ_REASON': 'category',
             'SENDING_LAB_PC': 'category',
             'SEQUENCING_LAB_PC': 'category',
             'lineage': 'category',
             'scorpio_call': 'category'
             }
)
#%%
df.rename(columns={
    'DATE_DRAW': 'date',
    'PROCESSING_DATE': 'processing_date',
    'SEQ_REASON': 'reason',
    'SENDING_LAB_PC': 'sending_pc',
    'SEQUENCING_LAB_PC': 'sequencing_pc',
    'lineage': 'lineage',
    'scorpio_call': 'scorpio'
    },
    inplace=True
)
df

License

The underlying files that I use as input are licensed by RKI under CC-BY 4.0, see more details here: https://github.com/robert-koch-institut/SARS-CoV-2-Sequenzdaten_aus_Deutschland#lizenz.

The software here is licensed under the "Unlicense". You can do with it whatever you want.

For the data, just cite the original source, no need to cite this repo since it's just a trivial join.