BETTER Synthetic Healthcare Dataset

This synthetic dataset release has been generated in the context of the Better Project, funded by the European Union’s Horizon program and the UK Research and Innovation Program.

The synthetic datasets mimic healthcare multimodal data (clinical and genomic) from different hospitals. It is intended to be reused for algorithm and pipeline testing for healthcare data management and analysis. The release includes healthcare data combining tabular data (in csv, tsv and xlsx format) and genomic resources in VCF format (Version 4.2 Specification). Three use cases are represented, with subfolders simulating data from distinct hospitals.

Content:

Use Case 1

condition: Hypotonia

synthetic datasets: VCF files and tabular data (xlsx, tsv)

description: For each of two hospitals, it is provided tabular data and VCF resources. Tabular data provides information on genomic, phenotypic, biological and clinical data (baseline and dynamic) according to domain rules. Regarding VCF resources, each dataset contains 300 patients, 50% of whom are affected by hypotonia and have a homozygote mutation overlapping the JUN gene region in the hg19 genome (chr1:59246463-59249719), and other random variants distributed throughout the genome.

Use Case 2

condition: Inherited Retinal Diseases

synthetic datasets: VCF files and tabular data (xlsx)

description: For each of two hospitals, it is provided tabular data and VCF resources. Tabular data provides information on electrophysiological and visual field data according to standard clinical eye examination protocols. The dynamic clinical data has been generated to mimic patients showing different clinical progression depending on the affected gene. Measures are generated for both eyes, which may present differences in progression. Regarding VCF resources, each dataset contains 200 patients, of which 50% have mutations overlapping gene ABCA4 and 50% with mutations overlapping gene RPGR. Details on included variants for each gene extracted from ClinVar can be found in the ANNEX below.

Use Case 3

condition: Autism Spectrum Disorder

synthetic datasets: tabular data (csv and tsv)

description: The generated dataset simulates patients with ASD, and provides information on demographic variables, neurodevelopmental and psychiatric comorbidities, family psychiatric history, bullying-related variables, abuse-related variables, sexual and gender-related descriptors, passive ideas of death, self-harm indicators and suicidal-intention-related features. Datasets have been generated with synthetic biases such as correlations and clusters to provide richness to the generated data. No VCF resources are included for this use case.

ANNEX (UC2)

The variants included in the VCF files for the ABCA4 gene are:

##fileformat=VCFv4.2

##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">

##FORMAT=<ID=GQ,Number=1,Type=Integer,Description="Genotype Quality">

##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">

##reference=hg19

#CHROM POS        ID    REF   ALT   QUAL  FILTER    FORMAT   SAMPLE_1

chr1 94471075         .           A         G         123      PASS               GT                   1/1

chr1 94476555         .           G         A         123      PASS               GT                   1/1

chr1 94480037         .           C         T          123      PASS               GT                   1/1

chr1 94487354         .           G         A         255      PASS               GT                   1/1

chr1 94495930         .           T          C         123      PASS               GT                   1/1

chr1 94544135         .           A         AC       255      PASS               GT                   1/1

chr1 94544234         .           T          C         255      PASS               GT                   1/1

chr1 94544276         .           G         A         255      PASS               GT                   1/1

chr1 94544805         .           T          TCA    255      PASS               GT                   1/1

chr1 94549029         .           A         G         255      PASS               GT                   1/1

chr1 94549083         .           T          C         123      PASS               GT                   1/1

chr1 94568491         .           A         G         255      PASS               GT                   1/1

chr1 94574053         .           C         A         255      PASS               GT                   1/1

chr1 94576968         .           T          C         255      PASS               GT                   1/1

chr1 94578548         .           T          C         123      PASS               GT                   1/1

The variants included in the VCF files for the RPGR gene are: 

##fileformat=VCFv4.2

##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">

##FORMAT=<ID=GQ,Number=1,Type=Integer,Description="Genotype Quality">

##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">

##reference=hg19

#CHROM POS        ID    REF   ALT   QUAL  FILTER    FORMAT   SAMPLE_1

chrX           38158290         .           C         T         123      PASS               GT                   1/1

chrX           38170131         .           C         G         123      PASS               GT                   1/1