Add robust parsing for temperature, time, and dosage for response curves

[tonywu1999]

Summary by CodeRabbit

• Chores

  • Updated package dependencies to include additional data manipulation libraries.

• Refactor

  • Streamlined internal dose-response fitting workflow for improved data handling and processing.
  • Updated UI label for dose-response selection from "Drug" to "Treatment" for clarity.

• Tests

  • Expanded test coverage for dose-response matrix construction and data editing functionality.

[coderabbitai]

📝 Walkthrough

Walkthrough

This PR refactors the dose-response workflow by replacing the convertGroupToNumericDose-dependent build_response_curve_matrix with a data-driven implementation. It introduces prepare_dose_response_fit helper function, adds utility for applying matrix edits, extends package imports (stringr, dplyr, tidyr), and updates corresponding tests and server logic.

Changes

Cohort / File(s)	Summary
Package Dependencies DESCRIPTION, NAMESPACE, R/MSstatsShiny.R	Added stringr to DESCRIPTION imports. Extended importFrom statements in NAMESPACE and MSstatsShiny.R to include dplyr functions (bind_rows, case_when, n, n_distinct, select, summarise, ungroup), stringr functions (str_detect, str_extract, str_extract_all, str_trim), and tidyr (pivot_wider).
Dose-Response Refactoring R/module-statmodel-server.R	Replaced build_response_curve_matrix implementation to derive GROUP column and is_control flag from group names instead of convertGroupToNumericDose; added prepare_dose_response_fit() function to standardize dose-response matrix schema; added update_matrix_from_edit() utility; updated render_group_comparison_plot_inputs and dose-response plot paths to use new helpers; changed UI label from "Select Drug" to "Select Treatment".
Utility Updates R/utils.R	Replaced MSstatsPrepareDoseResponseFit call in fitResponseCurves with prepare_dose_response_fit; updated importFrom directive to remove MSstatsPrepareDoseResponseFit dependency.
Tests tests/testthat/test-utils-statmodel-server.R	Added tests for build_response_curve_matrix with time-based and temperature-based scenarios; extended existing tests with stubbed showNotification; added tests for update_matrix_from_edit covering numeric and mixed-type data conversions.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

• feature(response-curve): Enable setting up metadata for response curves #133: Directly related as both modify build_response_curve_matrix flow; this PR replaces the convertGroupToNumericDose-based approach added there with a new parsing/reshape implementation.
• Configure editing setup matrix for dose response curves #143: Related through modifications to update_matrix_from_edit and editable matrix/contrast handling with accompanying tests.
• feat(response-curve): Enable modeling and plotting of dose response curves #139: Related through refactoring of dose-response code paths in both R/module-statmodel-server.R and R/utils.R, replacing MSstatsPrepareDoseResponseFit-based flow with new prepare_dose_response_fit helpers.

Suggested labels

Review effort 3/5

Suggested reviewers

• sszvetecz

Poem

🐰 A dose of refactoring hops into view,
Data-driven parsing, now fresh and new!
GroupNames decoded, matrices reshaped with care,
prepare_dose_response_fit—elegance in the air! ✨

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Title check	✅ Passed	The title accurately summarizes the main changes: adding robust parsing for temperature, time, and dosage parameters in response curve functionality, which is reflected in the build_response_curve_matrix refactoring and related updates throughout the codebase.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch fix-response-bugs

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-actions]

Failed to generate code suggestions for PR

[coderabbitai]

Actionable comments posted: 2

🤖 Fix all issues with AI agents

In `@R/module-statmodel-server.R`:
- Around line 261-287: The function prepare_dose_response_fit should validate
required columns before building cols_to_use: check that the expected response
column ("LogIntensities") and the expected protein column ("Protein") exist (or
explicitly allow a documented alternative), and if any are missing, stop() with
a clear error listing the missing columns rather than inserting NA and silently
dropping them; perform this validation after computing intervention_value but
before constructing cols_to_use/subset_df so downstream doseResponseFit gets a
predictable schema (refer to prepare_dose_response_fit, cols_to_use, subset_df,
and doseResponseFit).
- Around line 205-257: The build_response_curve_matrix function creates a
list-column measurements via str_extract_all which breaks downstream when groups
have multiple tokens; to fix, unnest measurements (use tidyr::unnest on the
measurements column) right after creating measurements so each row has a single
token, then extract numeric value and unit from that character column (use
str_extract on the unnested measurements) before pivot_wider (measurement_type,
value, unit); also remove or replace the direct UI call showNotification with a
returned warning value or a callback parameter so the data-processing function
remains UI-free and testable.

🧹 Nitpick comments (1)

tests/testthat/test-utils-statmodel-server.R (1)

380-420: Tests cover single-measurement groups well but lack multi-measurement coverage.

The tests validate the core scenarios (dose, time, temperature) but only for groups containing a single numeric+unit token. Consider adding a test for groups with multiple measurements (e.g., "Drug_10nM_24h") to verify the pivot_wider path and catch the list-column issue flagged in build_response_curve_matrix.