ModTect is a computational tool for detecting RNA modifications which disrupts base-pairing, from typical RNA-sequencing datasets. The method relies upon two features induced by base-pair-disrupting RNA modifications, (1) the multi-nucleotide-mismatch signal, and (2) the deletion signal. ModTect automatically extracts both signals from typical RNA-sequencing libraries, using a specialized statistical model, to determine possible RNA modification sites.
ModTect is written in Python and should not take >1-2 minutes to set-up on a system pre-installed with both Python2.7 and SAMtools.
ModTect was developed in a Linux environment. The following packages are also required for running:
- Python 2.7 (ModTect can also be executed using pypy for a ~2x speedup)
- SAMtools (version 0.1.19, or 1.3.1 and later)
- Download or clone the repository.
git clone https://github.com/ktan8/ModTect.git- Ensure that both Python 2.7 and SAMtools are installed and accessible in your $PATH. Note that python and samtools has to be callable via the following command.
Check if python 2.7 is accessible directly in your path:
python -VCheck if samtools is accessible directly in your path:
samtoolsNOTE: It is critical for samtools to be directly callable with the "samtools" command, as ModTect makes a direct call to samtools in your environment
The easiest way to run ModTect is as follows
python ModTect.py <bamfile_sorted_and_indexed> <Reference Genome> <chromosome> <position_start> <position_end>A full example is:
python ModTect_1_7_5_1.py ./sample/CML.MALAT1.bam Homo_sapiens_assembly19.fasta 11 65273620 65273640NOTE: The hg19 reference genome used for the sample dataset is the hg19 Broad variant (ftp://ftp.broadinstitute.org/pub/svtoolkit/reference_metadata_bundles/Homo_sapiens_assembly19_25Jan2015.tar.gz). In this reference genome, chromosomes are labelled "1,2,3,4,..." instead of "chr1,chr2,chr3,chr4,..."
A few optional arguments are also provided if you like to make modifications to it.
[--basesToTrimFromEdge BASESTOTRIMFROMEDGE]
[--scoreCutoff SCORECUTOFF]
[--minDepth MINDEPTH]
[--readlength READLENGTH]
[--mismatchPropFilter MISMATCHPROPFILTER]
[--deletionCountCutoff DELETIONCOUNTCUTOFF]
[--threads THREADS]
[--regionFile REGIONFILE]
[--label LABEL]
We have included a small sample dataset which you can use to test out ModTect. To run it, use the following command:
python ModTect_1_7_5.py ./sample/CML.MALAT1.bam Homo_sapiens_assembly19.fasta 11 65273620 65273640Please remember to point the "Homo_sapiens_assembly19.fasta" reference genome file to the path in your local machine.
You should see the following output:
chrom position reference_nt mono-alleleLogP bi-alleleLogP tri-alleleLogP tetra-alleleLogP variant_proportion depth deletionCount deletion_proportion refUpperCount refLowerCount altUpperCount altLowerCount refPlusProp altPlusProp A_count T_count G_count C_count a_count t_count g_count c_count has_reference_nt types_of_nt ModTect_score variant_readposn_median_fwd variant_readposn_median_rev median_abs_dev_fwd median_abs_dev_rev
11 65273630 A -183.627994199 -82.6673603595 -29.4593841182 -29.4593841182 0.734375 69 0.0724637681159 9 8 26 21 0.529411764706 0.553191489362 9 15 11 0 8 15 6 0 1 3 53.2079762413 44.5 38 21.5 22.0
We see the positive sites identified from this region, which has a modification score of "53.2079762413". This indicates that it is 10 ^ 53.2 more likely to be a tri-nucleotide site than a bi-nucleotide site.
These are the columns that you might find most useful:
- Modification score
- Deletion rate
- Variant proportion
- Depth
If you use ModTect for your work, please cite our study:
Kar-Tong Tan, Ling-Wen Ding, Chan-Shuo Wu, Daniel G. Tenen, Henry Yang. Repurposing RNA-Sequencing for Discovery of RNA Modifications in Clinical Cohorts. Science Advances (in press)
If you have any queries or spot any bugs, please contact me at Kar-Tong Tan (ktan@broadinstitute.org)