miRNome somatic mutations

Project containing Python scripts used to analyse somatic mutations in cancer (LUAD and LUSC) patients using somatic mutation data from TCGA

Python scripts may be reused for other data sources with input data prepared as somatic mutation data in TCGA (https://cancergenome.nih.gov/).

Results of all four algorithms available in TCGA database (muse, mutect2, somaticsniper, varscan2) were used.

For conditions to reuse of these scripts please refer to LICENSE file.

Pre-run preparation

All needed Python libraries are gathered in requirements.txt (Python 3 is needed).

Preparing ViennaRNA

Download ViennaRNA distribution from https://www.tbi.univie.ac.at/RNA/#download to chosen folder and install using instructions:

tar -zxvf ViennaRNA-2.4.11.tar.gz
cd ViennaRNA-2.4.11
./configure
make
sudo make install

Input data description

1) Coordinates

`\t` separated file without header with values like:

```chromosome\tstart\tstop\tsequence_name\n```

example:

```
chr1    17343   17456   hsa-mir-6859-1
chr1    30382   30483   hsa-mir-1302-2
chr1    632306  632428  hsa-mir-6723
```

2) Confidence file

Excel file with columns: `name, score, start, stop, Strand, id, confidence`
where name is `name` of miRNA, `score` is mirBase confidence score,
`start` and `stop` are coordinates, `Strand` is strand with `+` and `-` values,
`id` id mirBase_ID and `confidence` is miRBase confidence label with `High` and `Low` values.

Example row: `hsa-mir-1234, 0, 144400086, 144400165, -, MI0006324, Low`

Confidence file may be prepared with a script `prepare_confidence_file.py`
based on four files downloaded from miRBase `confidence.txt`,
`confidence_score.txt`,  `aliases.txt` and `mirna_chromosome_build`.

To run this script use:
```
python3 prepare_confidence_file.py ~/Documents/files_for_mirnaome/confidence.txt \
~/Documents/files_for_mirnaome/confidence_score.txt \
~/Documents/files_for_mirnaome/aliases.txt ~/Documents/files_for_mirnaome/mirna_chromosome_build.txt \
~/Documents/files_for_mirnaome/confidence_file.xlsx
```

3) mirgenedb file

Genomic coordinates from mirgenedb (http://mirgenedb.org/download) for human.
`hsa.gff` file

4) Cancer exons

Text file with names of exons that should be included in 
the first steps of the analysis (first mutations extraction
from vcf files, are included in coordinates file) but are not miRNA genes.
Single exon name in line.

Example:
```bash
EGFR_chr7.20
EGFR_chr7.21
EGFR_chr7.22
``` 

5) Localization file

Excel file with columns: `chrom, name, start, stop, orientation, based_on_coordinates, arm, type`
where name is `chrom` is chromosome id, `name` of miRNA localization, 
`start` and `stop` are coordinates, `orientation` is strand with `+` and `-` values,
`based_on_coordinates` states if localization is based on miRBase coordinates or
 structure prediction, `arm` is which arm of miRNA this sequence is on and `type` is type
 of localization within miRNA precursor.

Example row: `chr1, hsa-mir-6859-1-3p_post-seed, 17369, 17383, -, yes, 3p, post-seed`

Localization file may be prepared with a script `prepare_localization_file.py`
based on two files downloaded from miRBase `hairpin.fa`,
`hsa.gff` (here saved as `hsa.gff.txt` to differentiate from `hsa.gff` from mirgendb) and coordinates file. ViennaRNA is needed
(for installation see above). Important: add **absolute** path to Vienna package 

To run this script use:
```
python3 prepare_localization_file.py /home/<user>/Documents/ViennaRNA-2.4.11 ~/Documents/files_for_mirnaome/hairpin.fa \
~/Documents/files_for_mirnaome/hsa.gff3.txt\
~/Documents/files_for_mirnaome/new_coordinates_all_02.bed ~/Documents/output/Data_LUAD
```

6) (optional) Chromosome file

Tab-separated file with regions covered by NGS probes with columns 
`TargetID, Interval, Regions, Size, Databases, Coverage, HighCoverage, LowCoverage`

Example row: 
`HSA-LET-7A-1, chr9:96938239-96938318, 1, 80, CustomRegion, 100.0, 1, 0`

Output data description

1) temp folder

Contains merged vcf files if there were multiple samples available per single patient.

2) temp_reference folder

Temporary files created in create localization file script.

3) files_summary_count_per_patient.csv

File with information how many files there are per patient per algorithm. Sanity check: if the merging
of vcf files was successful, we should have only ones.

4) files_summary.csv

Files summary including user_id, file localization, sample id name and aliQ,
and algorithm used.

5) files_count_per_type.csv

Count of files per algorithm used.

6) not_unique_patients.csv

Patients for which we had multiple files (multiple samples) that were combined in a single vcf.

7) do_not_use.txt

Files that were replaced with merged vcf files (stored in temp folder)
that will not be used in next steps.

8) results_muse.csv, results_mutect2.csv, results_somaticsniper.csv, varscan2.csv

Mutations found in vcf files obtained in each of four algorithms.

9) results_muse_eval.csv, results_mutect2_eval.csv, results_somaticsniper_eval.csv, varscan2_eval.csv

Mutations found in vcf files obtained in each of four algorithms after additional evaluation methods
based on read counts, SSC, BQ and QSS.

10) all_mutations_filtered.csv

Mutations found by each of four algorithms concatenated.

11) all_mutations_filtered_merge_programs.csv

All mutations grouped to deduplicate mutations found by multiple algorithms.

12) all_mutations_filtered_mut_type_gene.csv

All mutations within miRNA genes with gene information, localization and mutation type.

13) complex.csv

Complex mutations are multiple mutations in single miRNA in single patient.
Column `complex` is `1` if mutation is treated as complex.

14) miRNA_per_chromosome.csv

How many miRNAs were mutated on single chromosome and how many mutations were found
in total on each chromosome.

15) occur.csv

How many total mutations, unique mutations and patients with mutation found per
gene.

16) distinct.csv

Unique mutations description with information how many patients had unique mutations.

17) distinct_with_loc.csv

Unique mutations description with localization within miRNA precursor.

18) (optional) mirnas_outside_probes.csv

Mutations in what miRNAs were detected in vcf files outside probes-defined regions.

19) (optional) high_confidence_mirnas_per_chrom.csv

miRNAs count per chromosome.

20) (optional) mirnas_per_chrom.csv

miRNAs mutated (found in vscf) count per chromosome.

21) (optional) patients_per_chrom.csv

Patients that had mutations in each chromosome.

How to use it

Example run is prepared in run_mirnaome.sh bash script.

To prepare confidence data run

python3 prepare_confidence_file.py ~/Documents/files_for_mirnaome/confidence.txt \
~/Documents/files_for_mirnaome/confidence_score.txt \
~/Documents/files_for_mirnaome/aliases.txt ~/Documents/files_for_mirnaome/mirna_chromosome_build.txt \
~/Documents/files_for_mirnaome/confidence_file.xlsx

Confidence file can be found in defined directory under defined filename.

To prepare localization data run

python3 prepare_localization_file.py ~/Documents/ViennaRNA-2.4.11 ~/Documents/files_for_mirnaome/hairpin.fa \
~/Documents/files_for_mirnaome/hsa.gff3.txt\
~/Documents/files_for_mirnaome/new_coordinates_all_02.bed ~/Documents/output/Data_LUAD

Localization file can be found in output folder.

To run analysis run

python3 run_mirnaome_analysis.py ~/dane/HNC/DATA_HNC ~/dane/HNC/RESULTS_HNC ./Reference/new_coordinates_all_02.bed \
./Reference/confidence.xlsx ./Reference/hsa.gff ./Reference/cancer_exons.txt

See additional features running

python3 run_mirnaome_analysis.py --help

To skip steps of the analysis (if first steps were already completed) use -s argument adding step from which script should start.

To include chromosome analysis use -c argument adding chromosome file path.

Authors

Martyna O. Urbanek-Trzeciak, Paulina Galka-Marciniak, Piotr Kozlowski

Institute of Bioorganic Chemistry, Polish Academy of Sciences, Noskowskiego 12/14, 61-704, Poznan, Poland

Citation

Somatic Mutations in miRNA Genes in Lung Cancer—Potential Functional Consequences of Non-Coding Sequence Variants

Paulina Galka-Marciniak¹, Martyna Olga Urbanek-Trzeciak¹, Paulina Maria Nawrocka¹, Agata Dutkiewicz¹, Maciej Giefing², Marzena Anna Lewandowska^3,4, and Piotr Kozlowski¹

¹ Institute of Bioorganic Chemistry, Polish Academy of Sciences, Poznan, Poland

² Institute of Human Genetics, Polish Academy of Sciences, Poznan, Poland

³ The F. Lukaszczyk Oncology Center, Department of Molecular Oncology and Genetics, Bydgoszcz, Poland

⁴ The Ludwik Rydygier Collegium Medicum, Department of Thoracic Surgery and Tumours, Nicolaus Copernicus University, Bydgoszcz, Poland

Cancers (MDPI) link to publication:

https://www.mdpi.com/2072-6694/11/6/793

Biorxiv link to preprint:

http://biorxiv.org/cgi/content/short/579011v1

Citation in MLA format

Galka-Marciniak, Paulina, et al. "Somatic Mutations in miRNA Genes in Lung Cancer—Potential Functional Consequences of Non-Coding Sequence Variants." Cancers 11.6 (2019): 793.

Contact

For any issues, please create a GitHub Issue.

Funding

This work was supported by research grants from the Polish National Science Centre [2016/22/A/NZ2/00184 (to P.K.) and 2015/17/N/NZ3/03629 (to M.O. U.-T.)] and the Polish Ministry of Science and Higher Education (support for young investigators to P.G.-M.)