A software toolkit for the interconversion of standard data models for phenotypic data
Documentation: https://cnag-biomedical-informatics.github.io/convert-pheno
Google Colab tutorial: https://colab.research.google.com/drive/1T6F3bLwfZyiYKD6fl1CIxs9vG068RHQ6?usp=sharing
CLI Source Code: https://github.com/cnag-biomedical-informatics/convert-pheno
CPAN Distribution: https://metacpan.org/pod/Convert::Pheno
Docker Hub Image: https://hub.docker.com/r/manuelrueda/convert-pheno/tags
Web App UI: https://convert-pheno.cnag.cat
convert-pheno - A script to interconvert common data models for phenotypic data
convert-pheno [-i input-type] <infile> [-o output-type] <outfile> [-options]
Arguments:
(input-type):
-ibff Beacon v2 Models ('individuals' JSON|YAML) file
-iomop OMOP-CDM CSV files or PostgreSQL dump
-ipxf Phenopacket v2 (JSON|YAML) file
-iredcap (experimental) REDCap (raw data) export CSV file
-icdisc (experimental) CDISC-ODM v1 XML file
-icsv (experimental) Raw data CSV
(Wish-list)
#-iopenehr openEHR
#-ifhir HL7/FHIR
(output-type):
-obff Beacon v2 Models ('individuals' JSON|YAML) file
-opxf Phenopacket v2 (JSON|YAML) file
(Wish-list)
#-oomop OMOP-CDM PostgreSQL dump
Compatible with -i(bff|pxf):
-ocsv Flatten data to CSV
-ojsonf Flatten data to 1D-JSON (or 1D-YAML if suffix is .yml|.yaml)
-ojsonld (experimental) JSON-LD (interoperable w/ RDF ecosystem; YAML-LD if suffix is .ymlld|.yamlld)
Options:
-exposures-file <file> CSV file with a list of 'concept_id' considered to be exposures (with -iomop)
-mapping-file <file> Fields mapping YAML (or JSON) file
-max-lines-sql <number> Maximum lines read per table from SQL dump [500]
-min-text-similarity-score <score> Minimum score for cosine similarity (or Sorensen-Dice coefficient) [0.8] (to be used with --search mixed)
-ohdsi-db Use Athena-OHDSI database (~2.2GB) with -iomop
-omop-tables <tables> OMOP-CDM tables to be processed. Tables <CONCEPT> and <PERSON> are always included.
-out-dir <directory> Output (existing) directory
-O Overwrite output file
-path-to-ohdsi-db <directory> Directory for the file <ohdsi.db>
-phl|print-hidden-labels Print original values (before DB mapping) of text fields <_labels>
-rcd|redcap-dictionary <file> REDCap data dictionary CSV file
-schema-file <file> Alternative JSON Schema for mapping file
-search <type> Type of search [>exact|mixed]
-svs|self-validate-schema Perform a self-validation of the JSON schema that defines mapping (requires IO::Socket::SSL)
-sep|separator <char> Delimiter character for CSV files [;] e.g., --sep $'\t'
-stream Enable incremental processing with -iomop and -obff [>no-stream|stream]
-sql2csv Print SQL TABLES (only valid with -iomop). Mutually exclusive with --stream
-test Does not print time-changing-events (useful for file-based cmp)
-text-similarity-method <method> The method used to compare values to DB [>cosine|dice]
-u|username <username> Set the username
Generic Options:
-debug <level> Print debugging level (from 1 to 5, being 5 max)
-help Brief help message
-log Save log file (JSON). If no argument is given then the log is named [convert-pheno-log.json]
-man Full documentation
-no-color Don't print colors to STDOUT [>color|no-color]
-v|verbose Verbosity on
-V|version Print Version
convert-pheno
is a command-line front-end to the CPAN's module Convert::Pheno.
A script that uses Convert::Pheno to interconvert common data models for phenotypic data
If you plan to only use the CLI, we recommend installing it via CPAN. See details below.
The script runs on command-line Linux and it has been tested on Debian/RedHat/MacOS based distributions (only showing commands for Debian's). Perl 5 is installed by default on Linux,
but we will install a few CPAN modules with cpanminus
.
First install system level dependencies:
sudo apt-get install cpanminus libbz2-dev zlib1g-dev libperl-dev libssl-dev
Now you have two choose between one of the 3 options below:
Option 1: System-level installation:
cpanm --notest --sudo Convert::Pheno
convert-pheno -h
Option 2: Install Convert-Pheno and the dependencies at ~/perl5
cpanm --local-lib=~/perl5 local::lib && eval $(perl -I ~/perl5/lib/perl5/ -Mlocal::lib)
cpanm --notest Convert::Pheno
convert-pheno --help
To ensure Perl recognizes your local modules every time you start a new terminal, you should type:
echo 'eval $(perl -I ~/perl5/lib/perl5/ -Mlocal::lib)' >> ~/.bashrc
Option 3: Install Convert-Pheno and the dependencies in a "virtual environment" (at local/
) . We'll be using the module Carton
for that:
mkdir local
cpanm --notest --local-lib=local/ Carton
echo "requires 'Convert::Pheno';" > cpanfile
export PATH=$PATH:local/bin; export PERL5LIB=$(pwd)/local/lib/perl5:$PERL5LIB
carton install
carton exec -- convert-pheno -help
Please follow these instructions.
git clone https://github.com/cnag-biomedical-informatics/convert-pheno.git
cd convert-pheno
Install system level dependencies:
sudo apt-get install cpanminus libbz2-dev zlib1g-dev libperl-dev libssl-dev
Now you have two choose between one of the 3 options below:
Option 1: Install dependencies (they're harmless to your system) as sudo
:
cpanm --notest --sudo --installdeps .
bin/convert-pheno --help
Option 2: Install the dependencies at ~/perl5
:
cpanm --local-lib=~/perl5 local::lib && eval $(perl -I ~/perl5/lib/perl5/ -Mlocal::lib)
cpanm --notest --installdeps .
bin/convert-pheno --help
To ensure Perl recognizes your local modules every time you start a new terminal, you should type:
echo 'eval $(perl -I ~/perl5/lib/perl5/ -Mlocal::lib)' >> ~/.bashrc
Option 3: Install the dependencies in a "virtual environment" (at local/
) . We'll be using the module Carton
for that:
mkdir local
cpanm --notest --local-lib=local/ Carton
export PATH=$PATH:local/bin; export PERL5LIB=$(pwd)/local/lib/perl5:$PERL5LIB
carton install
carton exec -- bin/convert-pheno -help
Download a docker image (latest version - amd64|x86-64) from Docker Hub by executing:
docker pull manuelrueda/convert-pheno:latest
docker image tag manuelrueda/convert-pheno:latest cnag/convert-pheno:latest
See additional instructions below.
Please download the Dockerfile
from the repo:
wget https://github.com/raw/cnag-biomedical-informatics/convert-pheno/main/Dockerfile
And then run:
docker buildx build -t cnag/convert-pheno:latest .
To run the container (detached) execute:
docker run -tid -e USERNAME=root --name convert-pheno cnag/convert-pheno:latest
To enter:
docker exec -ti convert-pheno bash
The command-line executable can be found at:
/usr/share/convert-pheno/bin/convert-pheno
The default container user is root
but you can also run the container as $UID=1000
(dockeruser
).
docker run --user 1000 -tid --name convert-pheno cnag/convert-pheno:latest
Alternatively, you can use make
to perform all the previous steps:
wget https://github.com/raw/cnag-biomedical-informatics/convert-pheno/main/Dockerfile
wget https://github.com/raw/cnag-biomedical-informatics/convert-pheno/main/makefile.docker
make -f makefile.docker install
make -f makefile.docker run
make -f makefile.docker enter
Docker containers are fully isolated. If you need the mount a volume to the container please use the following syntax (-v host:container
).
Find an example below (note that you need to change the paths to match yours):
docker run -tid --volume /media/mrueda/4TBT/data:/data --name convert-pheno-mount cnag/convert-pheno:latest
Then I will do something like this:
# First I create an alias to simplify invocation (from the host)
alias convert-pheno='docker exec -ti convert-pheno-mount /usr/share/convert-pheno/bin/convert-pheno'
# Now I use the alias to run the command (note that I use the flag --out-dir to specify the output directory)
convert-pheno -ibff /data/individuals.json -opxf pxf.json --out-dir /data
* Ideally a Debian-based distribution (Ubuntu or Mint), but any other (e.g., CentOs, OpenSuse, MacOS) should do as well.
(It should also work on macOS and Windows Server, but we are only providing information for Linux here)
* Perl 5 (>= 5.26 core; installed by default in most Linux distributions). Check the version with "perl -v".
* >= 4GB of RAM
* 1 core
* At least 16GB HDD
For executing convert-pheno you will need:
-
Input file(s):
A text file in one of the accepted formats. With
--iomop
I/O files can be gzipped. -
Optional:
Athena-OHDSI database
The database file is available at this link (~2.2GB). The database may be needed when using
-iomop
.Regardless if you're using the containerized or non-containerized version, the download procedure is the same. For CLI users, Google makes it difficult to use
wget
,curl
oraria2c
so we will use aPython
module instead:$ pip install gdown
And then run the following script
import gdown url = 'https://drive.google.com/uc?export=download&id=1-Ls1nmgxp-iW-8LkRIuNNdNytXa8kgNw' output = './ohdsi.db' gdown.download(url, output, quiet=False)
Once downloaded, you have two options:
a) Move the file
ohdsi.db
inside theshare/db/
directory.or
b) Use the option
--path-to-ohdsi-db
Examples:
$ bin/convert-pheno -ipxf phenopackets.json -obff individuals.json
$ $path/convert-pheno -ibff individuals.json -opxf phenopackets.yaml --out-dir my_out_dir
$ $path/convert-pheno -iredcap redcap.csv -opxf phenopackets.json --redcap-dictionary redcap_dict.csv --mapping-file mapping_file.yaml
$ $path/convert-pheno -iomop dump.sql -obff individuals.json
$ $path/convert-pheno -iomop dump.sql.gz -obff individuals.json.gz --stream -omop-tables measurement -verbose
$ $path/convert-pheno -cdisc cdisc_odm.xml -obff individuals.json --rcd redcap_dict.csv --mapping-file mapping_file.yaml --search mixed --min-text-similarity-score 0.6
$ $path/convert-pheno -iomop *csv -obff individuals.json -sep ','
$ carton exec -- $path/convert-pheno -ibff individuals.json -opxf phenopackets.json # If using Carton
* Error message: CSV_XS ERROR: 2023 - EIQ - QUO character not allowed @ rec 1 pos 21 field 1
Solution: Make sure you use the right character separator for your data with --sep <char>.
The script tries to guess it from the file extension, but sometimes extension and actual separator do not match.
When using REDCap as input, make sure that <--iredcap> and <--rcd> files use the same separator field.
The defauly value for the separator is ';'.
Example for tab separator in CLI.
--sep $'\t'
* Error message: Foo
Solution: Bar
The author requests that any published work that utilizes Convert-Pheno
includes a cite to the the following reference:
Rueda, M et al., (2024). Convert-Pheno: A software toolkit for the interconversion of standard data models for phenotypic data. Journal of Biomedical Informatics. DOI
Written by Manuel Rueda, PhD. Info about CNAG can be found at https://www.cnag.eu.
Copyright (C) 2022-2024, Manuel Rueda - CNAG.
This program is free software, you can redistribute it and/or modify it under the terms of the Artistic License version 2.0.