# Filter parameters for all-time global SARS-CoV-2 dataset

**URL:** <https://nextstrain.discourse.group/t/filter-parameters-for-all-time-global-sars-cov-2-dataset/1992>\
**Category:** General\
**Created:** [August 21, 2025, 12:36pm UTC](https://nextstrain.discourse.group/t/filter-parameters-for-all-time-global-sars-cov-2-dataset/1992 "2025-08-21T12:36:44Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![diegogotex](https://avatars.discourse-cdn.com/v4/letter/d/a587f6/32.png) [@diegogotex](https://nextstrain.discourse.group/u/diegogotex)\
**Post date:** [August 21, 2025, 12:36pm UTC](https://nextstrain.discourse.group/t/filter-parameters-for-all-time-global-sars-cov-2-dataset/1992/1 "2025-08-21T12:36:44Z")

</div>

Hi all,

I’m using the SARS-CoV-2 all-time global dataset ( [auspice](https://nextstrain.org/ncov/gisaid/global/all-time) ) to generate a small tree by adding some new samples.

Is there any documentation available about the filters that are applied to generate the sets for this tree?

Best regards,  
Diego

---

<div class="post-metadata">

**Author:** ![victorlin](https://yyz1.discourse-cdn.com/flex031/user_avatar/nextstrain.discourse.group/victorlin/32/306_2.png) [@victorlin](https://nextstrain.discourse.group/u/victorlin)\
**Post date:** [August 21, 2025, 9:13pm UTC](https://nextstrain.discourse.group/t/filter-parameters-for-all-time-global-sars-cov-2-dataset/1992/2 "2025-08-21T21:13:54Z")

</div>

Hi @diegogotex,

There is no specific documentation page, but the [workflow configuration file](https://github.com/nextstrain/ncov/blob/397b96226a837c4a3052e856f938b2cd4c2c71e8/nextstrain_profiles/nextstrain-gisaid/builds.yaml) has all the information. Here are the relevant entries for the `ncov/global/all-time` dataset.

```yaml
builds:
  global_all-time:
    subsampling_scheme: nextstrain_global_all_time

subsampling:
  # Custom subsampling logic for global region over all-time
  # 4320 total (expect ~3200)
  nextstrain_global_all_time:
    all:
      group_by: "country year month"
      group_by_weights: "defaults/population_weights.tsv"
      max_sequences: 4320

```

It aims to sample evenly over monthly intervals, with country-level population weighting used within each month’s sample, for a maximum of 4320 sequences in total.

– Victor

---

<div class="post-metadata">

**Author:** ![diegogotex](https://avatars.discourse-cdn.com/v4/letter/d/a587f6/32.png) [@diegogotex](https://nextstrain.discourse.group/u/diegogotex)\
**Post date:** [August 22, 2025, 11:10am UTC](https://nextstrain.discourse.group/t/filter-parameters-for-all-time-global-sars-cov-2-dataset/1992/3 "2025-08-22T11:10:03Z")

</div>

Thank you @victorlin!