A method to identify open chromatin regions, such as promotor areas, using a transposase.
- Sample type: Cells or nuclei
- Sample amount: 50,000 cells/nuclei per tube
- Sample collection and delivery: Samples should be delivered on ice at a time agreed upon with the responsible lab personnel at NGI.
- Mammalian cultured cells: Samples should be fresh and washed in PBS once, then pelleted.
- Plants: We only accept isolated nuclei which should preferably be obtained by using sucrose sedimentation.
- Nuclei can be frozen at –20˚C for up to two days.
- Tissues: We only accept extracted nuclei. Care should be taken to isolate nuclei with as little damage as possible.
Although the protocol allows for some margin of error the input amount should be kept the same across samples. Too few cells may cause the library prep chemistry to fail due to insufficient DNA input, while too many will result in a higher than optimal DNA-to-tagmentase ratio generating longer DNA fragments that will perform poorly on the sequencing machine. Therefore it is required to count the cells or nuclei before submission and turn in 50,000 cells per tube (in case of mammalian cells and plants).
For samples from sources with significantly higher or lower genome sizes (for example flies), the amount of cells need to be adjusted accordingly. Please ask your project manager for details.
How to evaluate the sample quality
We cannot check the correct amount of samples or the quality of your samples. You should make sure of the following:
- count the cells as accurately as possible
- don’t deliver more than 50.000 cells / sample
- make sure that your counted cells are alive – a high proportion of dead cells in the sample will reduce data quality
If you are not able to carry out these steps, or your samples are below the required thresholds, please get in touch.
What we do with your samples
When receiving cells from cell culture, we first spin down the cells and then resuspend the cells in a mild lysis buffer to generate a crude nuclei extract. The nuclei then are spun down and the transposition reaction is added directly.
The transposase attaches to DNA in accessible parts of the chromatin, fragmenting the DNA and at the same time adding partial Illumina adapters, thereby creating amplifiable DNA fragments.
After the transposition reaction, the samples undergo a preliminary amplification of 5 cycles after which the DNA content is determined by qPCR on an aliquot of the sample. The qPCR result determines any additional number of cycles. This process ensures that we only apply the actual number of PCR cycles required for your samples.
Library QC and sequencing
In this step, we evaluate the yield obtained and the size distribution of the libraries generated. We will inform you of the QC status of each sample. Once the libraries have passed this QC step, they are queued for sequencing to be carried out according to the setup stated in the agreement.
We have access to 96 (combinatorial) dual indexes so that is the highest level of multiplexing possible at the moment. Typically we recommend 50 million paired-end reads per library for human samples. The optimal depth likely depends on genome size and level of open chromatin, so for samples where these are expected to be very different from the typical sample, the sequencing depth may need to be adjusted.
Very generally speaking we have successfully been generating libraries that provide information on differential chromatin conformation. We expect more sequencing reads from euchromatic regions than from heterochromatic regions; areas with open chromatin conformation, for example the transcriptional start sites of active genes, are also expected to get more reads. Results will also depend on the characteristics of the sample and how they are sequenced:
- Paired end sequencing is required when information on nucleosome positioning is of interest.
- For general information on chromatin status changes, single end sequencing can be sufficient.
- Duplicate detection is significantly improved with paired-end sequencing.
- If too much DNA was provided, the fragment length distribution of the library will be shifted towards longer, high molecular weight fragments. We do not recommend the sequencing of such samples.
- If the nuclei were lysed before submission, we would expect higher background levels in the sequencing data.
The bioinformatics analysis that we provide for ATAC-seq is based on the nf-core/atacseq pipeline. It provides alignment, QC, filtering, peak calling, normalisation and differential accessibility analysis. Read about the pipeline outputs here: https://nf-co.re/atacseq/docs/output.
Due to limited experience in running this pipeline here, the analysis should be considered experimental in the same way as the lab work.