{%extends "phenotypes/base.html"%} {%from "flash_messages.html" import flash_all_messages%} {%from "macro-table-pagination.html" import table_pagination%} {%from "phenotypes/macro-display-pheno-dataset-card.html" import display_pheno_dataset_card%} {%block title%}Phenotypes{%endblock%} {%block pagetitle%}Phenotypes{%endblock%} {%block lvl4_breadcrumbs%}
We accept an extended form of the input files' format used with the R/qtl2 software as a single ZIP file
The files that are used for this feature are:
Other files within the bundle will be ignored, for this feature.
The following section will detail the expectations for each of the different file types within the uploaded ZIP file bundle for phenotypes:
There MUST be one, and only one file that acts as the control file. This file can be:
The control file is useful for defining things about the bundle such as:
sep: ','
). There can
only ever be one field separator and it MUST be the same
one for ALL files in the bundle.comment.char: '#'
). Any
line that starts with this character will be considered a comment line and
be ignored in its entirety.na.strings: 'NA'
). You
can specify more than one code to indicate missing values, e.g.
{…, "na.strings": ["NA", "N/A", "-"], …}
These files are the main data files. You must have at least one of these files in your bundle for it to be valid for this step.
The data is a matrix of individuals × phenotypes by default, as
below:
id,10001,10002,10003,10004,…
BXD1,61.400002,54.099998,483,49.799999,…
BXD2,49,50.099998,403,45.5,…
BXD5,62.5,53.299999,501,62.900002,…
BXD6,53.099998,55.099998,403,NA,…
⋮
If the pheno_transposed
value is set to True
,
then the data will be a phenotypes × individuals matrix as in the
example below:
id,BXD1,BXD2,BXD5,BXD6,…
10001,61.400002,49,62.5,53.099998,…
10002,54.099998,50.099998,53.299999,55.099998,…
10003,483,403,501,403,…
10004,49.799999,45.5,62.900002,NA,…
⋮
At least one phenotypes metadata file with the metadata values such as descriptions, PubMed Identifier, publication titles (if present), etc.
The data in this/these file(s) is a matrix of phenotypes × phenotypes-covariates. The first column is always the phenotype names/identifiers — same as in the R/qtl2 format.
phenocovar files should never be transposed!
This file MUST be present in the bundle, and have data for
the bundle to be considered valid by our system for this step.
In addition to that, the following are the fields that must be
present, and
have values, in the file before the file is considered valid:
The following optional fields can also be provided:
These files will be marked up in the control file with the
phenocovar
key, as in the examples below:
{
⋮,
"phenocovar": "your_covariates_file.csv",
⋮
}
{
⋮,
"phenocovar": [
"covariates_file_01.csv",
"covariates_file_01.csv",
⋮
],
⋮
}
⋮
phenocovar: your_covariates_file.csv
⋮
⋮
phenocovar:
- covariates_file_01.csv
- covariates_file_02.csv
- covariates_file_03.csv
…
⋮
These are extensions to the R/qtl2 standard, i.e. these types ofs file are not supported by the original R/qtl2 file format
We use these files to upload the standard errors (phenose) when the data file (pheno) is average data. In that case, the phenonum file(s) contains the number of individuals that were involved when computing the averages.
Both types of files are matrices of individuals × phenotypes by
default. Like the related pheno files, if
pheno_transposed: True
, then the file will be a matrix of
phenotypes × individuals.
Use the following resources for automated retrieval of certain data