Skip to content

miRTop/isomir_accuracy_meta_analysis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Draft

https://docs.google.com/document/d/1VQN_7j0omnzKNc566jfe39stZVhXsfG2feNEkaJ4cFc/edit

Goals

  • Optimize GFF format definition and usability
  • Detect methodology accuracy due to tools and some experimental step in the protocols.

Data

Tewari data

https://www.ncbi.nlm.nih.gov/pubmed/30010675

http://www.biorxiv.org/content/biorxiv/early/2017/05/17/113050.full.pdf

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=whipakmajrwprcv&acc=GSE94586

Narry Kim data

https://academic.oup.com/nar/article/47/5/2630/5271499

Carrie Wrigth data

https://www.biorxiv.org/content/10.1101/445437v3

DSRG data

Still to be published, another study to compare protocols using the mirxplor sample.

Fratta data

Evaluation of methodologies for microRNA biomarker detection by next generation sequencing https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6161688/ issue: wrong ID number to download data. Contacted the author to get the data.

To be added if we get the data

Systematic assessment of commercially available low-input miRNA library preparation kits https://www.biorxiv.org/content/10.1101/702456v1.full No data yet.

Processed data

Trimming was done with the smadann nexflow pipeline.

The following command was used for each study and type of data:

nextflow run mirtop/smadann --csv totrim.csv -c ../../om-profile.config --outdir trimmed -qs 10

Analysis was done with [bowtie] + [mirtop] in a [snakemake] file located in each study and data type.

snakemake -p -s run.snakefile

standard analysis of synthetic

Mirxplor reference was parsed to use only synthetic with an edit distance of 4 or more, and the alignments were filtered to keep only reads that mapped to those unique synthetic with a maximum of 4 changes. Code used for this is at analysis folder.

Data is available for anyone at aws mirtop space.

Currently contains: tewari, wrigth, kim and dsrg data.

biological samples

For human data we use miRBase22 to map all sequences. Same filtering step were used here.

Data is available for anyone at aws mirtop space.

Tools

  • bcbio smallRNA-seq pipeline + isomiRs - On charge Lorena Pantano
  • isomiR-SEA - On charge Gianvito Urgese
  • ChimiRa, miRge - On charge Marck Halushka
  • sRNAbench - On charge Michael Hackenberg
  • Prost - Thomas Desvignes
  • miRGe - Marc K. Halushka
  • (Add your tool here and person will do it)

Questions to address

  • Reproducibility of replicates
  • Reproducibility of protocols
  • Reproducibility of tools

Results

Updated report can be found here

Milestones:

Set up

  • Select random public data
  • Run with all the tools listed above
  • Put data in common space
  • Adapt output tools to GFF format

Random sample

Sample SRR5756178 is a whole blood small RNA-seq run from this manuscript https://academic.oup.com/nar/article/4080663 and is part of project PRJNA391912. It has ~ 2.8 million reads, of which ~2.6 million are miRNAs.

Synthetic data

Benchmark was done with synthetic isomiRs for one human miRNA, see results.

About

meta-analysis of public dataset to measure isomiR accuracy

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published