Update DENV Nextclade datasets with 2026 lineage definitions

[jamessiqueirap]

Description of proposed changes

This annual update incorporates 23 new lineages defined by the scientific committee across the four DENV serotypes, increasing the total from 218 to 241.

Serotype	Lineages through 2025	Lineages after the 2026 update	New lineages
DENV-1	70	75	5
DENV-2	82	94	12
DENV-3	37	41	4
DENV-4	29	31	2
Total	218	241	23

Test and representative sequences have also been updated to support the continued use of publicly available data. Test and representative datasets are now composed primarily of sequences available from Pathoplexus.

Checklist

• Assess whether the updated lineage assignments affect downstream workflows that depend on these datasets, including Nextstrain ingest workflows. Update affected workflows or open an issue describing any required changes.

[jamessiqueirap]

Hello @rneher @ivan-aksamentov @corneliusroemer, how are you guys doing?

I'm requesting to merge the dengue lineages dataset. I've included a short summary of the update in the description. If you notice any issues, please let me know.

Thanks!

[ivan-aksamentov]

Note

✨ AI is responding on behalf of Ivan

Review of PR #478, the 2026 update of community/v-gen-lab/dengue/denv1, denv2, denv3 and denv4, at head 7d43c1e against base ebb92c1.

Testing

• DENV-1
• DENV-2
• DENV-3
• DENV-4

What changed

Changes found in the diff, including the ones not mentioned in the description [click to expand]

The PR replaces tree.json and sequences.fasta and adds a CHANGELOG entry in each dataset. pathogen.json, reference.fasta, genome_annotation.gff3 and README.md are unchanged.

• New lineages: 5 in DENV-1, 12 in DENV-2, 4 in DENV-3 and 2 in DENV-4, for a total of 241, as in your description. Each tree sequence whose label changed moved from a parent lineage to one of its children (for example 2II_D.4 -> 2II_D.4.1). No lineage was removed or renamed
• Tree sequences: tips are now named by Pathoplexus accession (PP_...). When the names are mapped back to INSDC, 1,118 tips are the same sequences as before, 44 were removed and 182 were added
• Example sequences: replaced, mostly with Pathoplexus sequences. None of them are tips in the tree
• Changes not in the description or CHANGELOG:
  • Pathoplexus sequences with RESTRICTED data use terms (F1)
  • The outgroup tips are now labeled unassigned instead of Outgroup (F3)
  • The country attribute, the country coloring, the map data and the map panel were removed (F4)
  • A clade_membership color scale was added
  • A root_sequence was added. It is identical to reference.fasta
  • Gene annotations in tree.json now use prM and type gene, as in the GFF3. The base trees used M and CDS

Non-blocking issues

🟡 F1. RESTRICTED Pathoplexus sequences need a data use statement [click to expand]

On 2026-09-24, 36 of the Pathoplexus accessions in this PR had dataUseTerms: RESTRICTED (example query). They have no INSDC accession yet and become Open between 2027-03-17 and 2027-06-16.

Dataset	Tree tips	Examples	Submitting groups
DENV-1	13	3	USP Molecular Epidemiology Laboratory (13), DengSeq (3)
DENV-2	17	3	DengSeq (9), Gevop (8), Decode (3)

In these new lineages, every tree tip is RESTRICTED, so replacing them with Open sequences is not an option:

• 1V_D.1.4
• 2II_F.1.1.2.1
• 2II_F.1.1.2.2

The Pathoplexus terms allow sharing RESTRICTED data onward in web applications if the terms are clearly communicated and each sequence shows its Pathoplexus accession (Data Use Terms §4.2.4). The tip names already show the accessions. But the READMEs do not mention Pathoplexus, and meta.data_provenance in each tree.json lists only NCBI (denv1/tree.json#L7-L11). Users of the dataset cannot see that some of its sequences need the submitters' permission before use in a publication.

Suggestions:

• Add a short statement to each README.md, as in the nextstrain/orthoebolavirus/ebov dataset (README.md#L16-L18), for example:

  The reference tree and example sequences in this dataset come from [Pathoplexus](https://pathoplexus.org/) and include RESTRICTED sequences. It is not allowed to use these sequences in preprints or publications without permission of the submitters. See the [Pathoplexus Data Use Terms](https://pathoplexus.org/about/terms-of-use/data-use-terms) for details.

• Add Pathoplexus to meta.data_provenance in each tree.json:

  "data_provenance": [
    { "name": "Pathoplexus", "url": "https://pathoplexus.org/dengue" },
    { "name": "GenBank", "url": "https://www.ncbi.nlm.nih.gov/genbank/" }
  ]

• Optionally, add dataUseTerms and restrictedUntil as tip attributes, as the nextstrain/mpox trees do

• For the RESTRICTED example sequences, Open sequences from the same lineages would also work

🟡 F2. CHANGELOG does not list the new lineages [click to expand]

The new entry (denv1/CHANGELOG.md#L1-L4) uses the same text as the 2025 entry. Users read it to understand why their results changed, but it does not say which lineages are new. Sequences that were assigned to a parent lineage before can now get a child lineage. For example, the DENV-2 reference changes from 2V to 2V_F. The new lineages are not on dengue-lineages.org yet, so for now the CHANGELOG is the only place where users can find them.

Suggestions:

• List the new lineages for each serotype, for example:

  - New lineages: 1V_A.1, 1V_A.2, 1V_D.1.4, 1V_F.1, 1V_F.2

• Mention the other changes users will see: the Outgroup label (F3), the country coloring (F4), and the move to Pathoplexus accessions

• If the 2026 review has a public record, link to it

🟡 F3. README says sequences may be assigned as "Outgroup", but the trees no longer use this label [click to expand]

All READMEs say that sequences from other serotypes "may be assigned as 'Outgroup'" (denv1/README.md#L13). In the base trees, the outgroup tips (for example EF457905.1 in DENV-1 and KX274130.1 in DENV-2) had the label Outgroup. In the new trees, the same sequences have the label unassigned, so a query placed on an outgroup branch is now reported as unassigned.

Suggestions:

• If the change is intentional, update the README sentence and mention it in the CHANGELOG
• Otherwise, restore the Outgroup label on the outgroup tips

🔵 F4. Country data was removed, but the display settings still refer to it [click to expand]

You added the country attribute, the country coloring and the map during the first review of these datasets (PR #223). The new trees have none of these, but display_defaults.color_by is still "country", and filters still contains country and division (denv1/tree.json#L418). In the Nextclade Web tree view, users can no longer color by country, and the default coloring refers to a coloring that does not exist.

Suggestions:

• If the country data can come from the Pathoplexus metadata (geoLocCountry), restore it
• Otherwise, set color_by to clade_membership and remove country and division from filters

🔵 F5. "Only publicly available data" is not quite accurate [click to expand]

The CHANGELOG says that the tree sequences were updated "to retain only publicly available data". The trees still contain GISAID tips and one sequence without a public accession. EPI_ISL_17689858 is new in this PR:

• DENV-1: EPI_ISL_14908852
• DENV-2: EPI_ISL_17983090, EPI_ISL_17689875, EPI_ISL_17689858, EPI_ISL_17689850
• DENV-3: EPI_ISL_17689864, EPI_ISL_17689862, EHIE21409Y22

GISAID tips are also used in other datasets in this repository (for example nextstrain/flu), so only the wording needs a change.

Suggestions:

• Change the wording, for example: "Tree and example sequences now come mostly from Pathoplexus"

🔵 F6. Placement masks do not match the GFF3 UTR boundaries [click to expand]

This was already the case before this PR, but the trees are regenerated here, so it is easy to fix now. Placement masks are 0-based and end-exclusive, so the expected masks are [0, C start - 1) and [NS5 end, genome length):

Dataset	GFF3 C start	GFF3 NS5 end	Length	Expected	Actual
DENV-1	95	10,270	10,735	[0,94) [10270,10735)	[0,94) [10271,10735)
DENV-2	97	10,269	10,723	[0,96) [10269,10723)	[0,94) [10271,10735)
DENV-3	95	10,264	10,707	[0,94) [10264,10707)	[0,94) [10268,10707)
DENV-4	102	10,262	10,649	[0,101) [10262,10649)	[0,101) [10266,10649)

The DENV-2 masks are the same as the DENV-1 masks and end after the end of the DENV-2 genome. The effect on placement is small: 1 to 4 nt next to the coding region are not masked, and Nextclade does not fail on the out-of-range end.

Suggestions:

• Derive the masks from the GFF3 of each serotype, for example for DENV-2:

  "placement_mask_ranges": [
    { "begin": 0, "end": 96 },
    { "begin": 10269, "end": 10723 }
  ]

🔵 F7. The linked workflow repository does not contain the 2026 update yet [click to expand]

pathogen.json links to V-GEN-Lab/dengue-lineages-nextclade and the READMEs link to V-GEN-Lab/dengue-lineages-workflow. Both redirect to V-GEN-Lab/nextclade-datasets-workflow, where the last dengue commit is the 2025 update (b3afbd0). So the linked workflow does not yet show how the 2026 trees were built or how the sequences were selected from Pathoplexus.

Suggestions:

• Push the 2026 workflow changes, if they can be published
• Update the links in pathogen.json, the READMEs and meta.build_url to the current repository name

Questions

Questions about intent [click to expand]

• Q1. RESTRICTED sequences (F1): several of the submitting groups are your collaborators in the lineage consortium. Do you already have their agreement to include these sequences before the restriction ends? If so, the statement in F1 is all that is needed
• Q2. Outgroup label (F3): is unassigned for the outgroup tips intentional?
• Q3. Country data (F4): did you remove the country coloring on purpose, for example because of the move to Pathoplexus metadata?
• Q4. Workflow (F7): do you plan to publish the 2026 version of the workflow?

Clade distribution

Tip counts per lineage label. Sequences are matched by INSDC accession without the version suffix, so tips that were only renamed to Pathoplexus accessions count as unchanged. A sequence that moved to a child lineage counts as removed from the parent and added to the child. Only rows that changed are shown.

DENV-1: 387 -> 417 tips [click to expand]

Lineage	Base (ebb92c1)	Head (7d43c1e)	Removed	Added	Delta
1III	4	3	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
1III_A	6	7		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
1III_A.2	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
1III_A.4	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
1I_B	7	5	${\color{red}\text{-2}}$		${\color{red}\text{-2}}$
1I_B.1	4	6		${\color{green}\text{+2}}$	${\color{green}\text{+2}}$
1I_E.4	6	5	${\color{red}\text{-2}}$	${\color{green}\text{+1}}$	${\color{red}\text{-1}}$
1I_H.1	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
1I_K	12	9	${\color{red}\text{-3}}$		${\color{red}\text{-3}}$
1I_K.1.2	6	9	${\color{red}\text{-1}}$	${\color{green}\text{+4}}$	${\color{green}\text{+3}}$
1I_K.5	3	6		${\color{green}\text{+3}}$	${\color{green}\text{+3}}$
1V_A	6	6	${\color{red}\text{-2}}$	${\color{green}\text{+2}}$
1V_A.1	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
1V_A.2	0	5		${\color{green}\text{+5}}$	${\color{green}\text{+5}}$
1V_D.1	5	6		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
1V_D.1.2	6	5	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
1V_D.1.3	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
1V_D.1.4	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
1V_E	7	7	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
1V_E.2	6	5	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
1V_E.4	5	5	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
1V_F	6	6	${\color{red}\text{-3}}$	${\color{green}\text{+3}}$
1V_F.1	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
1V_F.2	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
Outgroup	1	0	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
unassigned	0	1		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
49 unchanged lineages	273	273
Total	387	417	${\color{red}\text{-23}}$	${\color{green}\text{+53}}$	${\color{green}\text{+30}}$

DENV-2: 438 -> 515 tips [click to expand]

Lineage	Base (ebb92c1)	Head (7d43c1e)	Removed	Added	Delta
2III	3	4	${\color{red}\text{-1}}$	${\color{green}\text{+2}}$	${\color{green}\text{+1}}$
2III_B	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
2III_C	4	3	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
2III_C.2	6	5	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
2III_C.3	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
2III_D.1	4	5	${\color{red}\text{-1}}$	${\color{green}\text{+2}}$	${\color{green}\text{+1}}$
2III_D.1.1	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
2III_D.1.2	6	6	${\color{red}\text{-2}}$	${\color{green}\text{+2}}$
2III_D.1.2.1	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
2III_D.1.2.2	0	7		${\color{green}\text{+7}}$	${\color{green}\text{+7}}$
2III_D.1.2.3	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
2III_D.1.3	5	6	${\color{red}\text{-1}}$	${\color{green}\text{+2}}$	${\color{green}\text{+1}}$
2III_D.2.1	5	5	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
2III_D.2.2	6	7		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
2III_D.3.1	6	7	${\color{red}\text{-1}}$	${\color{green}\text{+2}}$	${\color{green}\text{+1}}$
2III_D.3.2	6	7		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
2II_A	4	5		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
2II_A.1.1	4	3	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
2II_A.2.2	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
2II_D.4	6	2	${\color{red}\text{-4}}$		${\color{red}\text{-4}}$
2II_D.4.1	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
2II_D.4.2	0	7		${\color{green}\text{+7}}$	${\color{green}\text{+7}}$
2II_F.1.1	12	13	${\color{red}\text{-1}}$	${\color{green}\text{+2}}$	${\color{green}\text{+1}}$
2II_F.1.1.10	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
2II_F.1.1.2	6	10	${\color{red}\text{-1}}$	${\color{green}\text{+5}}$	${\color{green}\text{+4}}$
2II_F.1.1.2.1	0	5		${\color{green}\text{+5}}$	${\color{green}\text{+5}}$
2II_F.1.1.2.2	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
2II_F.1.1.2.3	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
2II_F.1.1.6	5	6		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
2II_F.1.1.9	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
2V	7	5	${\color{red}\text{-2}}$		${\color{red}\text{-2}}$
2V_A.1	5	4	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
2V_A.1.1	6	7		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
2V_F	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
Outgroup	1	0	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
unassigned	0	1		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
59 unchanged lineages	313	313
Total	438	515	${\color{red}\text{-23}}$	${\color{green}\text{+100}}$	${\color{green}\text{+77}}$

DENV-3: 189 -> 209 tips [click to expand]

Lineage	Base (ebb92c1)	Head (7d43c1e)	Removed	Added	Delta
3III_B.1	6	5	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
3III_B.4	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
3III_B.5	0	5		${\color{green}\text{+5}}$	${\color{green}\text{+5}}$
3III_C.1	6	7		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
3III_C.2.1	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
3III_C.2.2	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
3II_A.1	6	7	${\color{red}\text{-3}}$	${\color{green}\text{+4}}$	${\color{green}\text{+1}}$
3II_B	5	4	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
3I_C	1	2		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
3I_C.2	6	2	${\color{red}\text{-4}}$		${\color{red}\text{-4}}$
3I_C.2.1	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
3I_C.2.2	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
3I_C.2.3	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
Outgroup	1	0	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
unassigned	0	1		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
28 unchanged lineages	140	140
Total	189	209	${\color{red}\text{-13}}$	${\color{green}\text{+33}}$	${\color{green}\text{+20}}$

DENV-4: 149 -> 160 tips [click to expand]

Lineage	Base (ebb92c1)	Head (7d43c1e)	Removed	Added	Delta
4II_A.1	4	3	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
4II_A.1.3	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
4II_B	6	4	${\color{red}\text{-2}}$		${\color{red}\text{-2}}$
4II_B.2	5	6		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
4II_B.3	0	6		${\color{green}\text{+6}}$	${\color{green}\text{+6}}$
4I_A.1.2	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
4I_B	3	4		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
4I_B.1	6	6	${\color{red}\text{-1}}$	${\color{green}\text{+1}}$
Outgroup	1	0	${\color{red}\text{-1}}$		${\color{red}\text{-1}}$
unassigned	0	1		${\color{green}\text{+1}}$	${\color{green}\text{+1}}$
22 unchanged lineages	118	118
Total	149	160	${\color{red}\text{-6}}$	${\color{green}\text{+17}}$	${\color{green}\text{+11}}$

Validation summary

Validation checks [click to expand]

Lineages

• New lineage labels per serotype match the description
• No lineage removed; each changed label on an existing tree sequence moves from a parent to its child
• Every new lineage has at least 5 tips in the tree
• Fail: label Outgroup replaced by unassigned, with no README update -- see F3

GFF3 annotation and reference (unchanged in this PR)

• All gene lengths divisible by 3, C starts with ATG in all references
• No ambiguous bases in the references; headers match pathogen.json
• tree.json gene annotations match the GFF3 coordinates and names
• root_sequence.nuc is identical to reference.fasta
• Fail: placement masks differ from the GFF3 UTR boundaries -- see F6

Tree metadata

• All tip names carry a date; tip dates span 1953 to 2026
• Fail: data_provenance lists only NCBI -- see F1
• Fail: color_by and filters refer to the removed country data -- see F4

Data sources

• Pathoplexus accessions: 1,488 Open, 36 RESTRICTED -- see F1
• GISAID tips and one tip without a public accession -- see F5

Build

• scripts/rebuild succeeds for these datasets; the generated files match the source files

Nextclade CLI 3.23.0

• References: QC good, 0 substitutions; lineages 1IV_B, 2V_F, 3III_A, 4II_B
• Private mutations of the references: 0, 2, 37 and 63 (base: 0, 2, 35 and 64)
• Examples good: 62/70, 83/92, 35/38 and 25/31. All bad results come from missing data (1,110 to 3,288 N) in partial genomes
• Compared with the base dataset on the same examples, QC status is the same for every sequence, and each changed lineage call moves to a more specific lineage (see N1)

Notes

Click to expand

• N1. On the example sequences, most changed lineage calls go from a parent to a new child. Two DENV-3 examples move from 3III_B.2 to the new sibling 3III_B.5, and two examples move to children that already existed (1I_B.1 and 2II_F.1.1.6)
• N2. The new clade_membership color scale has no color for the labels that occur only on internal nodes: 1I_K.1, 1V_D, 2II_A.2 and 4II
• N3. OK040058.2 (PP_0064LPM) is labeled DENV-1 in GenBank and on Pathoplexus, but it is in the DENV-4 tree with only 27 mutations on its branch (lineage 4II_A.1.1.2). Your placement is probably correct, and the upstream serotype label is probably wrong. It may be worth reporting to the submitters
• N4. In tree.json, the seqid of the gene annotations is a relative path from a local checkout (../nextclade_data/data_output/.../genome_annotation.gff3). This has no effect on Nextclade

[ivan-aksamentov]

@jamessiqueirap Thank you for the update! As an engineer I have no technical complaints. I asked my AI to screen these changes further. If you agree with any points, please feel free to reply and/or update the dataset. I will let my human scientist colleagues to review the scientific parts.

[rneher]

@jamessiqueirap this looks good. But I since this is using data from Pathoplexus, we need to make sure that we link back to the original source, and the DataUseTerms. For example data, it would be best if this contained only Open data.

One way to do this would be to include this provenance info into the tree such that the tool tips look like this:

To this end, you need to add URL fields to the metadata. We have a script that does this for many workflows and it looks like this:

https://github.com/nextstrain/rsv/blob/master/ingest/bin/curate-urls.py

Then the auspice config should contain the metadata columns to use:

https://github.com/nextstrain/rsv/blob/master/config/auspice_config.json#L96-L100

[rneher]

The nextstrain docs page has some more detailed instructions:
https://docs.nextstrain.org/en/latest/guides/pathoplexus.html

[jamessiqueirap]

Hi guys @ivan-aksamentov @rneher, thanks for the feedback! I’m working through all the points and will resubmit it here shortly. Thanks!

[jamessiqueirap]

Thanks @rneher and @ivan-aksamentov for the detailed review. I’ve revised the DENV update to address the points raised:

1. Pathoplexus data use and provenance (F1): The dataset READMEs now include the applicable Pathoplexus data-use notice. The trees include both Pathoplexus and GenBank provenance, and tip metadata now includes accession links, data-use terms, and restriction dates. The assembly step excludes Pathoplexus example sequences marked RESTRICTED, while restricted sequences required in the reference trees retain their recorded data-use terms.

2. CHANGELOG (F2): The new entries now list the lineages added for each serotype and describe the user-visible changes to outgroup assignments, country metadata, and the transition to Pathoplexus identifiers.

3. Outgroup wording (F3): The READMEs and CHANGELOGs now reflect the unassigned result shown by the updated trees.

4. Country metadata (F4): Country information from the metadata snapshot has been restored to the trees, together with country coloring and map settings.

5. Sequence-source wording (F5): The documentation no longer states that all sequences are publicly available and instead describes the tree and example sequences as coming mostly from Pathoplexus.

6. Placement masks (F6): The workflow has been revised so that placement masks are derived independently for each serotype from its GFF3 coding boundaries and reference genome length. I’m completing a final reproducibility check of this step before requesting another review.

7. Published workflow and links (F7): I’m still working on updating the workflow repository. Since lineage updates are developed collaboratively, I’ve made some adjustments to the workflow used to update the Nextclade lineage datasets.

Richard, I also followed your [suggestion regarding source and data-use links](#478 (comment)): accession and data-use fields are now prepared as linked tree-tip metadata.

I’m also considering changing the institution name in the directory structure. Would there be any issue with doing that?

In any case, considering the addition of the new lineages included in this update, I was thinking of releasing this functional update first and making those structural changes afterwards. What do you think?

[rneher]

@jamessiqueirap thanks so much for updating the datasets and bringing them inline with the data use terms. We try to have an eye on this to maintain the trust people have in the data sharing mechanism on pathoplexus

regarding the name/directory change... this is in principle possible. But it might break links to the datasets that other people have bookmarked or hard-coded. It might be possible to include the old links as "short-cuts" into the pathogen json. But I am not sure whether such short cuts are allowed for names that were previously full path. @ivan-aksamentov would know more. One could also envision to move the dengue datasets into a top-level group like the nextstrain group. e.g. dengue-lineages/DENV1 etc

[jamessiqueirap]

Thanks, @rneher ! That makes sense, and preserving existing links is definitely something I’d want to avoid breaking.
For now, I think it would be better to keep this structural change separate from the current lineage update, so we can get the updated datasets released first and address the directory reorganization afterwards.