📦 Subsampled "Toy" Datasets Instant Downloads (< 15 MB) GIAB • SEQC2 • 1000 Genomes

Curated Practice Datasets & Benchmark Sandboxes

Don't wait for massive 50 GB downloads just to test a pipeline or debug a script. Access curated, lightweight practice datasets (100k paired-end FASTQs, downsampled BAMs, synthetic cancer VCFs, 500-cell single-cell matrices, and mock microbiomes) designed to run in seconds on any laptop or CI/CD test runner.

🚀 Subsampled Practice Datasets (Under 15 MB)

Laptop & Cloud Ready

FASTQ (Paired-End) Size: ~12.4 MB

Human chr21 Exome Mini-FASTQs (100k Reads)

Subsampled from Genome in a Bottle reference standard HG002 (NA24385) targeting chromosome 21 coding exons. Perfect for validating BWA-MEM2 alignment, FastQC quality checking, and GATK4 variant calling in under 2 minutes.

curl -sLO https://omics.co.in/downloads/datasets/toy_chr21_R1.fq.gz
curl -sLO https://omics.co.in/downloads/datasets/toy_chr21_R2.fq.gz

MD5 R1: 4f28b7e91a0c4519d8329b3f9e8a71d2

BAM • CRAM Size: ~14.8 MB

Mini Tumor-Normal BAM Pair (TP53 & BRCA1)

Coordinate-sorted, indexed paired BAM files covering TP53 (chr17:7.5-7.7Mb) and BRCA1 (chr17:43.0-43.2Mb). Includes read duplicates and true low-frequency somatic mutations (VAF ~ 18%) for Mutect2 benchmarking.

curl -sLO https://omics.co.in/downloads/datasets/mini_tumor.bam
curl -sLO https://omics.co.in/downloads/datasets/mini_normal.bam

Includes .bai indices • Aligned to GRCh38 / hg38

VCF 4.3 Size: ~1.8 MB

Synthetic Oncology VCF with Actionable Hotspots

Annotated somatic callset with known drug-actionable variants (EGFR L858R, BRAF V600E, KRAS G12D, PIK3CA H1047R) along with benign polymorphisms and sequencing artifacts for practicing clinical curation.

curl -sLO https://omics.co.in/downloads/datasets/synthetic_cancer.vcf.gz
bcftools index -t synthetic_cancer.vcf.gz

Contains 1,450 variants • Multi-sample FORMAT tags

10x Genomics Size: ~4.2 MB

10x Chromium 500-Cell PBMC Matrix

Filtered feature-barcode sparse matrix representing 500 human peripheral blood mononuclear cells (T cells, B cells, Monocytes, NK cells). Loadable in Seurat v5 or Scanpy in under 15 seconds.

curl -sLO https://omics.co.in/downloads/datasets/pbmc500_matrix.tar.gz
tar -xzf pbmc500_matrix.tar.gz

Contains matrix.mtx.gz, barcodes.tsv.gz, features.tsv.gz

16S rRNA Mock Size: ~5.6 MB

20-Strain Staggered Bacterial Mock Community

Illumina MiSeq 2×250 bp FASTQ pair sequencing the V4 region of a 20-strain staggered mock microbial community. Excellent for testing DADA2 ASV error rate modeling and QIIME 2 pipelines.

curl -sLO https://omics.co.in/downloads/datasets/mock16s_R1.fq.gz
curl -sLO https://omics.co.in/downloads/datasets/mock16s_R2.fq.gz

BEI Resources MSA-1003 standard reference

🏛️ Gold-Standard Public Genomic Reference Benchmarks

Truth Sets

Benchmark Project	Key Sample Accession	Technology & Coverage	Primary Purpose	Repository Link
Genome in a Bottle (GIAB)	HG002 (NA24385 - Son)	Illumina NovaSeq 300x, PacBio HiFi 50x, ONT Ultra-long	High-confidence Small Variant & Structural Variant Truth Benchmark	NIST GIAB FTP →
FDA SEQC2 Consortium	HCC1395 / HCC1395BL	Matched Tumor-Normal WGS across 6 sequencing centers	Somatic Mutation Calling & RNA-seq Quantification Accuracy	NCBI BioProject →
1000 Genomes Project	NA12878 (CEPH Mother)	Illumina 30x Deep Coverage (NYGC)	Population Allele Frequencies & Haplotype Phasing Benchmark	IGSR Portal →

🛠️ How to Generate Your Own Toy Datasets (CLI Guide)

When testing large pipelines, create downsampled versions of your raw files using these deterministic commands:

# 1. Subsample exactly 100k paired FASTQ reads with seqtk:

seqtk sample -s100 read1.fq.gz 100000 | gzip > mini_R1.fq.gz

seqtk sample -s100 read2.fq.gz 100000 | gzip > mini_R2.fq.gz

# 2. Subsample 5% of reads from a BAM file:

samtools view -s 42.05 -b full.bam > mini.bam

# '42' is random seed, '.05' is 5% fraction

# 3. Extract specific chromosome interval from VCF:

bcftools view -r chr17:7500000-7700000 full.vcf.gz -Oz -o tp53.vcf.gz