Summary of Data Requirements
To ensure optimal performance of the CytoTerra platform, datasets should contain 150–250 million Hi-C read pairs, with at least:
- 100 million high-quality read pairs
- ≥20% same-strand read pairs
Reads must be submitted as separate forward (R1) and reverse (R2) FASTQ files in gzipped format. The steps below will help you assess and prepare your data for submission.
1. Downsampling
If you’re unsure how many read pairs you have, use one of the commands below to count lines in each FASTQ file. Run the command on both R1 and R2 files to ensure they are not corrupted and contain matching read counts (a slight difference is okay).
For a gzipped FASTQ:
zcat sample_R1.fastq.gz | wc -l
For an unzipped FASTQ:
wc -l sample_R1.fastq
Since each read occupies 4 lines, divide the total line count by 4 to determine the number of reads. If the dataset exceeds 250 million read pairs, downsampling is required.
Downsampling with BBTools
1. Download and extract BBTools:
wget https://sourceforge.net/projects/bbmap/files/latest/download -O bbtools.tar.gz
tar -xvzf bbtools.tar.gz
2. Add to PATH (adjust install path as needed):
echo 'export PATH=~/bbtools_install/bbmap:$PATH' >> ~/.bashrc
source ~/.bashrc
3. Run the reformat.sh tool to downsample:
Determine the samplerate by dividing your desired target (e.g., 150 million) by your current number of read pairs.
The example below uses samplerate=0.2; adjust this value accordingly.
reformat.sh \
in1=example_R1.fastq.gz \
in2=example_R2.fastq.gz \
out1=example_downsampled_R1.fastq.gz \
out2=example_downsampled_R2.fastq.gz \
samplerate=0.2
2. Hi-C QC
After downsampling, please run our Hi-C QC tool to assess data quality. We recommend that you use the HG38 assembly.
3. Data Manifest
Please download and complete the Data Manifest.
4. Data Upload
Once your data and manifest are ready:
- Provide Phase Genomics with your data manifest to receive sFTP credentials
- Use those credentials to upload your files to our secure server by following our sFTP Guide