Convert output of cross-product bridge to be used in downstream functions #472

dtopouza · 2024-11-13T21:48:57Z

Title: Convert output of cross-product bridge to be used in downstream functions

Problem: Output of olink_normalization_product is not suitable for input to downstream OA functions

Solution: Added a function to format cross-product bridging output for downstream analysis, and added an argument to olink_normalization to toggle it.

Key Features:

olink_normalization_product_format function
- Input is olink_normalization_product result
- Removes internal and external controls
- Replaces SampleID with the concatenation of SampleID and Project to make unique sample IDs for downstream analysis
- Removes non-bridgeable assays
- Replaces the NPX values of the non-reference project by the Median Centered or QS Normalized NPX, according to the Bridging Recommendation.
- Replaces OlinkID by the concatenation of the Explore HT and Explore 3072 OlinkIDs to record the OlinkIDs from both projects.
- Removes BridgingRecommendation, MedianCenteredNPX, QSNormalizedNPX, OlinkID_E3072 columns
olink_normalization function
- Adds the format argument, a Boolean that controls whether the normalized dataset will be formatted for input to downstream analysis. Only applicable for cross-product bridge normalization.

Checklist

Make sure you are requesting to pull a feature/bugfix branch (right side). This should not be main or develop.
Make sure you are make a pull request against either main or develop (left side). (Requesting to main should be reserved for bugfixes and new releases)
Add or update unit tests (if applicable)
Check your code with any unit tests (Run devtools::check() locally)
Add neccessary documentation (if applicable)

Type of changes

What type of changes does your code introduce?
Put an x in the boxes that apply

Bugfix (non-breaking change which fixes an issue) (link the issue on the right)
New feature (non-breaking change which adds functionality)
Breaking change (fix or feature that would cause existing functionality to not work as expected)
Code style update (formatting, renaming)
Documentation Update
Other (explain)

…product output

…k_normalization_product_format is run. Add argument to olink_normalization to toggle whether formatting is applied.

…k-Proteomics/OlinkRPackage into develop_product_bridging_downstream

OlinkAnalyze/R/olink_normalization.R

kristynchin-olink · 2024-11-19T16:37:55Z

OlinkAnalyze/R/olink_normalization_product.R

+    dplyr::mutate(SampleID = paste0(.data[["SampleID"]],
+                                     "_",
+                                     .data[["Project"]])) |>
+    dplyr::filter(.data[["BridgingRecommendation"]] != "NotBridgeable") |>


Removes NotBridgeable datapoints as expected per PR notes.

Question: do we want to remove these? Or can we keep the NPX as is in these cases and keep these lines in?

The NotBridgeable assays don't have a recommendation for MedianCenteredNPX or QSNormalizedNPX, I'm not sure how to include them.

Kathy suggested to keep them with the 3k NPX values and 3k OlinkID and the HT ones with just the HT OlinkID.

I could not get this to work - downstream tests like t.test would throw an error, since the not bridgeable assay HT and 3072 OlinkIDs are only present in some samples and not others.

Resolved, NotBridgeable and NotOverlapping assays are retained

OlinkAnalyze/R/olink_normalization_product.R

kristynchin-olink · 2024-11-19T16:40:31Z

OlinkAnalyze/R/olink_normalization_product.R

+      .data[["BridgingRecommendation"]] == "QuantileSmoothing" ~
+        .data[["QSNormalizedNPX"]],
+      .default = .data[["NPX"]])) |>
+    dplyr::filter(.data[["AssayType"]] == "assay") |>


Removes internal controls as appropriate. Looks good.

kristynchin-olink · 2024-11-19T16:40:45Z

OlinkAnalyze/R/olink_normalization_product.R

+
+  ### Keep the data following BridgingRecommendation
+  df_format <- df |>
+    dplyr::filter(.data[["SampleType"]] == "SAMPLE") |> # Remove controls


Removes external controls as appropriate. Looks good.

Do we want to leave in SCs? Might be a more complicated conversation that Kathy, Amrita, and I have touched base on. May be added functionality for a later release.

SCs have NA values for their QSNormalizedNPX, not sure how to include them. We would have to change the QS normalization function as well - may need to consult Kathy/Klev

Kathy suggested that we keep SCs out for now, and potentially include them in a future release once we update the QC normalization function

Noted, the current code reflects that external controls, including SCs, are removed from the formatted DF. Looks good.

OlinkAnalyze/R/olink_normalization_product.R

kristynchin-olink · 2024-11-19T16:42:38Z

OlinkAnalyze/R/olink_normalization_product.R

+    dplyr::mutate(OlinkID = paste0(.data[["OlinkID"]],
+                                   "_",
+                                   .data[["OlinkID_E3072"]])) |>
+    dplyr::select(!c(.data[["BridgingRecommendation"]],


Bridging rec, median normalized NPX, QS normalized NPX, and 3K OID columns removed as appropriate. HT OID column replaced with concatenated HT OID and 3K OID. Looks good.

kristynchin-olink · 2024-11-19T16:52:52Z

OlinkAnalyze/tests/testthat/test-olink_normalization_product.R

All tests in this file (old and new) passed on my end - great job!

All tests (old and new) have once again passed, looks good.

…nto develop_product_bridging_downstream

…d NotOverlapping assays with their original OlinkIDs and NPX values.

…nto develop_product_bridging_downstream

kristynchin-olink · 2025-01-24T15:49:19Z

OlinkAnalyze/R/olink_normalization_product.R

                     .data[["QSNormalizedNPX"]],
                     .data[["OlinkID_E3072"]]))# Remove extra columns

-  return(df_format)
+  df_full <- rbind(df_format,


Suggestion / optional change: does it make sense to arrange the combined dataframe by Project or Project-SampleID? Not sure if this is super important to implement or something we could push to a later release.

kristynchin-olink · 2025-01-24T15:50:25Z

OlinkAnalyze/R/olink_normalization_product.R

+#' OlinkID by the concatenation of the Explore HT and Explore 3072 OlinkIDs to
+#' record the OlinkIDs from both projects for bridgeable assays. Assays that are
+#' NotBridgeable or NotOverlapping retain their original non-reference OlinkIDs
+#' and NPX values. Replaces SampleID with the concatenation of SampleID and


I like that we reincorporate the NotOverlapping datapoints! I can make a mention of this in the 3K-HT vignette. No action needed on your end.

Confirmed that NotOverlapping assays are incorporated into the formatted DF as expected, looks good.

dtopouza and others added 6 commits November 12, 2024 16:52

Add initial function and tests for formatting of olink_normalization_…

e23df0d

…product output

Remove control samples and concatenate SampleID and Project when olin…

a4ec970

…k_normalization_product_format is run. Add argument to olink_normalization to toggle whether formatting is applied.

Edit function description

418fa83

Document

b3df23b

Fix function description

fd5b993

Merge branch 'develop_product_bridging_downstream' of github.com:Olin…

789e542

…k-Proteomics/OlinkRPackage into develop_product_bridging_downstream

kristynchin-olink reviewed Nov 19, 2024

View reviewed changes

OlinkAnalyze/R/olink_normalization.R Outdated Show resolved Hide resolved

kristynchin-olink reviewed Nov 19, 2024

View reviewed changes

dtopouza and others added 7 commits November 19, 2024 13:51

Merge branch 'develop' of github.com:Olink-Proteomics/OlinkRPackage i…

aff86b6

…nto develop_product_bridging_downstream

Update olink_normalization_product_format to include NotBridgeable an…

f0335c6

…d NotOverlapping assays with their original OlinkIDs and NPX values.

Merge branch 'develop' of github.com:Olink-Proteomics/OlinkRPackage i…

7be56c0

…nto develop_product_bridging_downstream

Update get_example_data function and calls with new file location

05e954b

fix test example data interaction

c6ffb9b

added missing selective call for dpylr stars_with

0899058

Fix dplyr package call and confirm devtools::checks() pass

a2f78b0

kristynchin-olink reviewed Jan 24, 2025

View reviewed changes

kristynchin-olink approved these changes Jan 24, 2025

View reviewed changes

Order output by Project and SampleID

f162903

dtopouza merged commit 2f285ae into develop Jan 24, 2025
14 of 16 checks passed

kristynchin-olink mentioned this pull request Jan 24, 2025

update Description and news for v4.1 #492

Merged

11 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Convert output of cross-product bridge to be used in downstream functions #472

Convert output of cross-product bridge to be used in downstream functions #472

dtopouza commented Nov 13, 2024

kristynchin-olink Nov 19, 2024

dtopouza Nov 19, 2024

dtopouza Nov 19, 2024

dtopouza Nov 29, 2024

dtopouza Jan 24, 2025

kristynchin-olink Nov 19, 2024

kristynchin-olink Nov 19, 2024

kristynchin-olink Nov 19, 2024

dtopouza Nov 19, 2024

dtopouza Nov 19, 2024

kristynchin-olink Jan 24, 2025

kristynchin-olink Nov 19, 2024

kristynchin-olink Nov 19, 2024

kristynchin-olink Jan 24, 2025

kristynchin-olink Jan 24, 2025

dtopouza Jan 24, 2025

kristynchin-olink Jan 24, 2025

kristynchin-olink Jan 24, 2025

Convert output of cross-product bridge to be used in downstream functions #472

Convert output of cross-product bridge to be used in downstream functions #472

Conversation

dtopouza commented Nov 13, 2024

Title: Convert output of cross-product bridge to be used in downstream functions

Checklist

Type of changes

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment