Data

Tools

Indexes

Applications

Experiments

Open Source Science

crumblr

Count ratio uncertainty modeling base linear regression

https://github.com/diseaseneurogenomics/crumblr

Science Score: 23.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ○
    CITATION.cff file
  • ○
    codemeta.json file
  • ○
    .zenodo.json file
  • ✓
    DOI references
    Found 5 DOI reference(s) in README
  • ✓
    Academic publication links
    Links to: biorxiv.org, springer.com
  • ○
    Committers with academic emails
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (11.3%) to scientific vocabulary
Last synced: 11 months ago · JSON representation

Repository

Count ratio uncertainty modeling base linear regression

Basic Info
Statistics
  • Stars: 6
  • Watchers: 0
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Fork of GabrielHoffman/crumblr
Created about 2 years ago · Last pushed 12 months ago

https://github.com/DiseaseNeuroGenomics/crumblr/blob/main/ 




### Count ratio uncertainty modeling based linear regression 






	
The `crumblr` package enables analysis of count ratio data using precision-weighted linear (mixed) models, PCA and clustering.  `crumblr`'s fast, normal approximation of transformed count data from a Dirichlet-multinomial model allows use of standard workflows to analyze count ratio data while modeling heteroskedasticity.

![](man/figures/Figure_crumblr_workflow.png)

__Preprint:__ Hoffman and Roussos. 2025. Fast, flexible analysis of differences in cellular composition with crumblr. [biorxiv](https://www.biorxiv.org/content/10.1101/2025.01.29.635498v1)


### Details
Analysis of count ratio data (i.e. fractions) requires special consideration since data is non-normal, heteroskedastic, and spans a low rank space.  While counts can be considered directly using Poisson, negative binomial, or Dirichlet-multinomial models for simple regression applications, these can be problematic since they 1) can be very computationally expensive, 2) can produce poorly calibrated hypothesis tests, and 3) are challenging to extend to other applications.  The widely used centered log-ratio (CLR) transform from [compositional data analysis](https://link.springer.com/book/10.1007/978-3-642-36809-7) makes count ratio data more normal and enables use the linear models, and other standard methods.  

Yet CLR-transformed data is still highly heteroskedastic: the precision of measurements varies widely. This important factor is not considered by existing methods. 

`crumblr` uses a fast asymptotic normal approximation of CLR-transformed counts from a Dirichlet-multinomial distribution to model the sampling variance of the transformed counts.  `crumblr` enables incorporating the sampling variance as precision weights to linear (mixed) models in order to increase power and control the false positive rate.  `crumblr` also uses a variance stabilizing transform (vst) based on the precision weights to improve performance of PCA and clustering.      





### Install
```r
# 1) Make sure Bioconductor is installed
if (!require("BiocManager", quietly = TRUE)) {
  install.packages("BiocManager")
}

# 2) Install crumblr and dependencies
BiocManager::install('DiseaseNeurogenomics/crumblr')
```

### Introduction to compositional data analysis

- Brief intro for bioinformatics [Quinn, et al. 2018](https://doi.org/10.1093/bioinformatics/bty175)
- Book for analysis in R  [van den Boogaart and Tolosana-Delgado, 2013](https://link.springer.com/book/10.1007/978-3-642-36809-7)

Owner

  • Name: Center for Disease Neurogenomics @ Mount Sinai
  • Login: DiseaseNeuroGenomics
  • Kind: organization
  • Location: United States of America

Open source software, documentation and applications for analysis of functional genomics datasets

GitHub Events

Total
  • Watch event: 5
  • Delete event: 1
  • Push event: 13
  • Pull request event: 2
  • Create event: 1
Last Year
  • Watch event: 5
  • Delete event: 1
  • Push event: 13
  • Pull request event: 2
  • Create event: 1

Committers

Last synced: 11 months ago

All Time
  • Total Commits: 94
  • Total Committers: 2
  • Avg Commits per committer: 47.0
  • Development Distribution Score (DDS): 0.021
Past Year
  • Commits: 31
  • Committers: 2
  • Avg Commits per committer: 15.5
  • Development Distribution Score (DDS): 0.065
Top Committers
Name Email Commits
Gabriel Hoffman g****n@g****m 92
lahuuki l****i@g****m 2

Issues and Pull Requests

Last synced: 11 months ago

All Time
  • Total issues: 0
  • Total pull requests: 1
  • Average time to close issues: N/A
  • Average time to close pull requests: less than a minute
  • Total issue authors: 0
  • Total pull request authors: 1
  • Average comments per issue: 0
  • Average comments per pull request: 0.0
  • Merged pull requests: 1
  • Bot issues: 0
  • Bot pull requests: 0
Past Year
  • Issues: 0
  • Pull requests: 1
  • Average time to close issues: N/A
  • Average time to close pull requests: less than a minute
  • Issue authors: 0
  • Pull request authors: 1
  • Average comments per issue: 0
  • Average comments per pull request: 0.0
  • Merged pull requests: 1
  • Bot issues: 0
  • Bot pull requests: 0
View more stats
Top Authors
Issue Authors
Pull Request Authors
  • GabrielHoffman (1)
Top Labels
Issue Labels
Pull Request Labels

Packages

  • Total packages: 1
  • Total downloads:
    • bioconductor 1,263 total
  • Total dependent packages: 0
  • Total dependent repositories: 0
  • Total versions: 1
  • Total maintainers: 1
bioconductor.org: crumblr

Count ratio uncertainty modeling base linear regression

  • Versions: 1
  • Dependent Packages: 0
  • Dependent Repositories: 0
  • Downloads: 1,263 Total
Rankings
Dependent repos count: 0.0%
Dependent packages count: 30.3%
Average: 41.2%
Downloads: 93.2%
Maintainers (1)
gabriel.hoffman@mssm.edu
Last synced: 12 months ago