Skip to content

Milab 6935 input types rework - #57

Open
mchernys wants to merge 23 commits into
mainfrom
MILAB-6935_input-types-rework
Open

mchernys wants to merge 23 commits into
mainfrom
MILAB-6935_input-types-rework

Conversation

@mchernys

@mchernys mchernys commented Oct 22, 2026 •

Copy link
Copy Markdown
Contributor

RetriggerView in GreptileConfidence Score: 5/5

The PR appears safe to merge; the remaining documentation issue is non-blocking.

Fix All in Claude CodeFindings

  1. P2 Count retention is overstated ▶
Fix with agent prompt
### Issue 1
README.md:undefined-57
`README.md` promises that recognized formats keep every count, but immunoSEQ builds one `read-count`: it uses `templates` when filled and otherwise `reads`. It also sets `hasUMIs` to false. Users may expect both source counts downstream, but this preset keeps only one. Document immunoSEQ’s one-count exception here and in the overview.

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

Summary

Reworks format imports, chain pairing, and import statistics.

  • V(D)J imports recognize more export layouts and sequence regions.
  • Import statistics show how many rows each filter removes.
  • Single-cell selections include complete receptor pairs.
  • Custom imports map source headers to the fields users need.

Important touched types:

  • BlockArgs: Settings sent to the workflow. Single-cell selections now include partner chains; primaryCountType is sent only for Custom imports.
  • BlockData: Saved settings and table state. Existing table layouts reset once to use the new column defaults.
  • BareSetMapping: Maps named records to chain sequences and optional genes. New blocks remove mappings for undeclared chains.
  • AssemblingFeature: Names the sequence region used to group clonotypes. MiXCR now supports an explicit choice alongside AIRR.
  • MappableColumn: Describes a Custom mapping control. Its new optional required field lets either CDR3 sequence satisfy the sequence requirement.

Diagram

%%{init: {'theme': 'neutral'}}%%
flowchart LR
  A[Source tables] --> B[Infer columns]
  B --> C[Apply format rules]
  C --> D[Count discarded rows]
  C --> E{Import mode}
  E --> F[Bulk clonotypes]
  E --> G[Pair cell chains]
  E --> H[Annotate named records]
  F --> I[Dataset and statistics]
  G --> I
  H --> I
Loading

Reviews (3) · Last reviewed commit: "MILAB-6935: restore the README count par..." · Reviewed by Greptile

Comment thread workflow/src/import-immunoSeq.tpl.tengo Resolved
@mchernys

Copy link
Copy Markdown
Contributor Author

@greptileai

Comment thread README.md Outdated
### What is the secondary count type for?

Some pipelines report both read counts and UMI counts. Configure the primary count type you want abundances based on, and carry the other as a secondary count so both remain available downstream.
Some pipelines report both read counts and UMI counts. In Custom mode, set the primary count type you want abundances based on, and carry the other as a secondary count so both remain available downstream. Recognized formats carry every count they have and base abundances on UMIs when present.

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Count retention is overstated

README.md promises that recognized formats keep every count, but immunoSEQ builds one read-count: it uses templates when filled and otherwise reads. It also sets hasUMIs to false. Users may expect both source counts downstream, but this preset keeps only one. Document immunoSEQ’s one-count exception here and in the overview.

Knowledge Base Used: Source format and column inference

Prompt To Fix With AI
This is a comment left during a code review.
Path: README.md
Line: 57

Comment:
**Count retention is overstated**

`README.md` promises that recognized formats keep every count, but immunoSEQ builds one `read-count`: it uses `templates` when filled and otherwise `reads`. It also sets `hasUMIs` to false. Users may expect both source counts downstream, but this preset keeps only one. Document immunoSEQ’s one-count exception here and in the overview.

**Knowledge Base Used:** [Source format and column inference](https://app.greptile.com/milaboratories/-/custom-context/knowledge-base/platforma-open/import-vdj-data/-/docs/source-format-and-column-inference.md)

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!

Fix in Claude Code

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

README reverted to previous state in c1284da

@mchernys

Copy link
Copy Markdown
Contributor Author

@greptileai

@julenmendieta julenmendieta left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I shared a document with some issues I found, and that need to be adressed

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants