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/1T6F3bLwfZyiYKD6fl1CIxs9vG068RHQ

CLI Source Code: https://github.com/cnag-biomedical-informatics/convert-pheno

CPAN Distribution: https://metacpan.org/pod/Convert::Pheno

Web App UI Source Code: https://github.com/cnag-biomedical-informatics/convert-pheno-ui

Docker Hub Image: https://hub.docker.com/r/manuelrueda/convert-pheno/tags

NAME

convert-pheno - A script to interconvert common data models for phenotypic data

SYNOPSIS

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 number of lines read 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

DESCRIPTION

convert-pheno is a command-line front-end to the CPAN's module Convert::Pheno.

SUMMARY

A script that uses Convert::Pheno to interconvert common data models for phenotypic data

INSTALLATION

If you plan to only use the CLI, we recommend installing it via CPAN. See details below.

Containerized

Method 1: From Docker Hub

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.

Method 2: With Dockerfile

Please download the Dockerfile from the repo:

wget https://raw.githubusercontent.com/cnag-biomedical-informatics/convert-pheno/main/Dockerfile

And then run:

docker buildx build -t cnag/convert-pheno:latest .

Additional instructions for Methods 1 and 2

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://raw.githubusercontent.com/cnag-biomedical-informatics/convert-pheno/main/Dockerfile
wget https://raw.githubusercontent.com/cnag-biomedical-informatics/convert-pheno/main/makefile.docker
make -f makefile.docker install
make -f makefile.docker run
make -f makefile.docker enter

Mounting volumes

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

Non containerized

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.

From Github

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

From CPAN

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

System requirements

* Ideally a Debian-based distribution (Ubuntu or Mint), but any other (e.g., CentOs, OpenSuse, MacOS) should do as well.
* 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

HOW TO RUN CONVERT-PHENO

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. In Linux you can use wget, curl or aria2c:

    $ wget 'https://drive.google.com/uc?export=download&id=1-Ls1nmgxp-iW-8LkRIuNNdNytXa8kgNw&confirm=t' -O ohdsi.db --no-check-certificate
    or
    $ curl -L 'https://drive.google.com/uc?export=download&id=1-Ls1nmgxp-iW-8LkRIuNNdNytXa8kgNw&confirm=t' > ohdsi.db
    or
    $ aria2c -x2 'https://drive.google.com/uc?export=download&id=1-Ls1nmgxp-iW-8LkRIuNNdNytXa8kgNw&confirm=t' -o ohdsi.db
  

  (you can install wget, curl or aria2c inside the container by typing sudo apt install wget, sudo apt install curl or sudo apt install aria2.

  Once downloaded, you have two options:

  a) Move the file ohdsi.db inside the share/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

COMMON ERRORS AND SOLUTIONS

* 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. 
            The script tries to guess it from the file extension (e.g. comma for csv), 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.
  Example for tab separator in CLI.
   --sep  $'\t' 

* Error message: Foo
  Solution: Bar

CITATION

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

AUTHOR

Written by Manuel Rueda, PhD. Info about CNAG can be found at https://www.cnag.eu.

COPYRIGHT AND LICENSE

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.