Skip to content

Latest commit

 

History

History
35 lines (30 loc) · 2.75 KB

metadata_info.md

File metadata and controls

35 lines (30 loc) · 2.75 KB
Raw

Description of metadata

#TODO- update to match cellxgene requirements https://github.com/chanzuckerberg/single-cell-curation/tree/main/schema/4.1.0

Source/citation info:

  • source.label - shorthand for the publication from which this data was taken (first author's last name plus year final manuscript was published)
  • sample - sample ID, specific to each lane of a 10x Chromium run (string of characters with no spaces, no periods, and no hyphens/dashes)
  • description - Brief description of the sample. Not used by any pipeline.
  • include - Whether or not to include the sample in the analysis (True/False)
  • source - This is an abbreviate citation for each publication (Author et al, Journal, year)

Biological metadata & experimental conditions

  • species - species from which the sample was derived (Genus species)
  • sex - ("M"/"F")
  • age - measured in... days? weeks? months? #TODO
  • tissue - #TODO
  • subtissue - #TODO

Technical info:

  • chemistry - This column specifies what chemistry/method was used to prepare the data (10x Genomics Chromium v1/v2/v3/v3.1, or other methods). Note For lots of samples where chemistry was not stated beyond "10x Genomics", I used the length of R1 to determine if the sample was v2 or v3 (v3 and v3.1 use the same R1 length, so specifying isn't as important; please let me know if you find an error though!). See resources/chemistry_sheet.csv for details on how this is used in alignment (align_snake).
  • sequencer - type of sequencer used

Accession info:

  • file.format - This column contains the file format in which authors originally uploaded their data. This is used by the align_snake pipeline to download and reformat each file so that they may be re-aligned to a common reference. Options:
    • "fastq"
    • "bam"
    • "aws" - file is downloaded w/ wget and renamed to match the accession info #TODO
    • "local" - option for files that you have downloaded manually, or files that aren't public (yet...) #TODO
  • GSE.accession - Gene Expression Omnibus (GEO) project accession number (GSE#######)
  • GSM.accession - GEO sample accession number (GSM#######). Note this ID is used for pooling fastq's (SRR #'s) from samples for which authors uploaded multiple files
  • SRR.accession - This column contains the "SRR#########" IDs for each run. These IDs are important because they are used to download the raw sequencing data for each sample (see align_snake). For samples which have more than one run, the SRR numbers are delimited with a semi-colon (;). ("SRR#########" or "ERR########")
  • SAMN.accession - INSERT_DESCRIPTION_HERE
  • other.accession - INSERT_DESCRIPTION_HERE
  • file.link - link that can be used to directly download files with wget