Repository navigation
ROADMAP 2025 #57
Copy link
Copy link
Open
Labels
Description
Activity
Spring 2025
Cardinal 3.10 (Bioconductor 3.21 - expected April)
This will be our 10th anniversary edition! We can be ambitious. Among other new features, this should finally be the release where we introduce a GUI.
- Graphical user interface
- Based on shiny
- Visualization only (for now)
- Maybe preview pre-processing?
iMSI()?MSIView()?msiView()?
- Ion mobility (experimental)
- Support for timsTOF via TIMSCONVERT to imzML
- Add
readTIMSData()function - Read mobility arrays with cvParam "MS:1003006"
- Load data as
TIMSImagingExperiment - Add
peakProcess()method forTIMSImagingExperiment - Pre-release on Github in February/March?
- Co-registration (experimental)
- Native support for simple linear and affine transformations
- Add
coregistermethod based onmatter::warp2_trans() - Aim to be good enough for visualization, but not necessarily annotation
- More advanced methods coming later in CardinalNN
CardinalNN 0.99 (under development)
This will be our companion package for any neural network-based methods requiring Python.
- Prototype R package
- Define Basilisk virtual environments
- Methods for serializing core objects to Python
- Need to figure out how to map to Python classes
- May need to pull data into memory for
matter->numpy.ndarray - Eventually we'll want native
matterobjects in Python
- Bindings for msireg?
- Need to coordinate with Sai
- Bindings for mzclustering?
- Khoury Research Apprenticeship project
Reacted by AMAwatermelon- Graphical user interface
- pinned this issue
on Dec 19, 2024 None of the above happened because US politics overwhelmed the author.
Let's try again.
Spring 2026
Cardinal 3.14 (Bioconductor 3.23 - expected April)
This update will be focused on usability and cleaning up the public API.
- Refactor core classes
- Decouple
MSImagingArraysandMSImagingExperiment - Use
MSImagingArraysfor processing spectra - Use
MSImagingExperimentfor downstream analysis
- Decouple
- Refactor spectral processing
- Harmonize processing function implementations
- Refactor
process() - Refactor
peakProcess() - Remove unnecessary generic functions
- Refactor results storage
- Refactor imzML import/export
- Refactor
pixels()andfeatures()
CardinalStats (under development)
Statistical learning methods will be moved to a new package.
- Prototype CardinalStats package
- Implement nearest neighbor methods (BiocNeighbors)
- Implement dimension reduction methods (BiocSingular)
- Implement clustering methods (bluster)
- Refactor core classes
This issue serves as a tentative outline of development planned for 2025.
IF YOU ARE A USER, DO NOT MAKE RESEARCH PLANS BASED ON THIS TIMELINE.