A pipeline to analyse PASP data
Reference
This repository contains the pipeline and raw results described in the manuscript:
- Botond Sipos, Adrian M. Stütz, Greg Slodkowicz, Tim Massingham, Jan Korbel, Nick Goldman: PASP - a whole-transcriptome poly(A) tail length determination assay for the Illumina platform.
Click here for more details on the wetlab experiment.
Using the pipeline
The pipeline can be used by invoking the following make targets:
- Fetch raw data from ENA:
make fetch - Generate transcriptome from SGD annotation:
make transcriptome - Align and parse reads:
make parse - Test for differential polyadenylation:
make testormake lsf_test - Parse spike-in reads (
make parse_spikeins) and build "error model" (make error_model) - Filter test results by G-tail coverage:
make gtail_cov_filter - Plot correlation between technical replicates:
make gtail_tech_corr - Plot and cluster tail length distributions:
make classify_tail_dists - Correlate thresholded tail lengths with PASTA and PAL-seq:
make corr_with_studies
Index of selected raw results
Illumina reads
Sequencing data are available in the ArrayExpress database under accession number E-MTAB-2456.
Alignment
Parsing alignments
Quantifying tail length slippage using spike-in standards
- Tail run lengths until the first 1-5 non-A bases in reads mapped to spike-in poly(A) tracts PDF
Testing differences between wild type and mutant tail runs
| Comparison | Test log | Test report | Results |
|---|---|---|---|
| WT1 vs. MUT1 | TXT | CSV | |
| WT2 vs. MUT2 | TXT | CSV |
Tail run distributions from all transcripts with G-tail coverage > 1000
Cross-study correlation
| WT1 vs. MUT1 | http://bit.ly/PAT-seq-TEST_WT1_vs_MUT1_log | http://bit.ly/PAT-seq-TEST_WT1_vs_MUT1_pdf | http://bit.ly/PAT-seq-TEST_WT1_vs_MUT1_trs_tab | | WT2 vs. MUT2 | http://bit.ly/PAT-seq-TEST_WT2_vs_MUT2_log | http://bit.ly/PAT-seq-TEST_WT2_vs_MUT2_pdf | http://bit.ly/PAT-seq-TEST_WT2_vs_MUT2_trs_tab |
Tail run distributions from all transcripts with G-tail coverage > 1000
- WT1: http://bit.ly/PAT-seq-CLS_WT1_pdf
- WT2: http://bit.ly/PAT-seq_CLS_WT2_pdf
- MUT1: http://bit.ly/PAT-seq-CLS_MUT1_pdf
- MUT2: http://bit.ly/PAT-seq-CLS_MUT2_pdf
Cross-study correlation
- PAL_total vs. WT1: http://bit.ly/PAT_seq_PAL_total_vs_WT1_pdf
- PAL_total vs. WT2: http://bit.ly/PAT-seq_PAL_total_vs_WT2_pdf
- PAL_total vs. PASTA: http://bit.ly/PAT-seq-PAL_total_vs_PASTA_pdf
Dependencies
- Platform LSF
- Python 2.x
- numpy >= 1.6.2
- matplotlib >= 1.1.0
- scipy >= 0.10.1
- biopython >= 1.60
- Bowtie2 >= 2.1.0
- samtools >= 0.1.19+
- wget
Using the analysis tools
The analysis tool can be found under patsy/:
patsy-align - classify read pairs and align them using Bowtie2
usage: patsy-align [-h] -1 fq1 -2 fq2 -f ref [-o outdir] [-s stats_pickle]
[-l gtail_sig] [-G gtag_min] [-N max_N] [-I min_fsize]
[-X max_fsize] [-p nr_threads] [-r report]
Align PASP reads using Bowtie2 (1.1).
optional arguments:
-h, --help show this help message and exit
-1 fq1 First FASTQ file.
-2 fq2 Second FASTQ file.
-f ref Reference fasta.
-o outdir Output directory.
-s stats_pickle Stats pickle file.
-l gtail_sig Portion of read start/end used for G-tail classification
(14).
-G gtag_min Minimum G-tag length(3).
-N max_N Maximum number of Ns in the first -l bases (6).
-I min_fsize Minimum fragment size (0).
-X max_fsize Maximum fragment size (500).
-p nr_threads Number of threads to use (1).
-r report Report PDF.
patsy-parse - parse classified and aligned PASP read pairs
usage: patsy-parse [-h] -g gtail_sam -n nvtr_sam -d dataset_id -f ref
[-l gtail_sig] [-G gtag_min] [-N max_N] [-e err_tol]
[-o out_pickle] [-i tr_list] [-q min_q] [-r report] [-t]
Parse classified and aligned PASP read pairs (1.1).
optional arguments:
-h, --help show this help message and exit
-g gtail_sam SAM file containing G-tail alignments.
-n nvtr_sam SAM file containing NVTR alignments.
-d dataset_id Dataset identifier.
-f ref Reference fasta.
-l gtail_sig Portion of read start/end used for G-tail classification.
-G gtag_min Minimum G-tag length(3).
-N max_N Maximum number of Ns in the first -l bases (6).
-e err_tol Number of errors tolerated in the tail.
-o out_pickle Output pickle file.
-i tr_list List of transcripts considered.
-q min_q Mapping quality treshold (30).
-r report Report PDF.
-t Plot per-transcript coverage reports.
patsy-test - test for differential polyadenylation in PASP data
usage: patsy-test [-h] -a [a_pickles [a_pickles ...]] -na a_name -b
[b_pickles [b_pickles ...]] -nb b_name [-i lrt_list]
[-P lik_penalty] [-M min_size_U] [-s sig_level]
[-op out_pickle] [-ot out_trs] [-og out_glob]
[-otr out_runs_prefix] [-orr out_rep_prefix] [-r report]
[-t]
Test for differential polyadenylation in PASP data (1.1).
optional arguments:
-h, --help show this help message and exit
-a [a_pickles [a_pickles ...]]
Parsed read pickles - group A.
-na a_name Name of data group A.
-b [b_pickles [b_pickles ...]]
Parsed read pickles - group B.
-nb b_name Name of data group B.
-i lrt_list Transcripts to be tested with anchors LRT.
-P lik_penalty Log-likelihood penalty for data points outside valid
range.
-M min_size_U Minimum sample size when performing Mann-Whitney U
test (30).
-s sig_level Significance level.
-op out_pickle Output pickle file.
-ot out_trs Output tabular file: transcript properties.
-og out_glob Output tabular file: global results.
-otr out_runs_prefix Output tabular file: tail runs prefix.
-orr out_rep_prefix Output tabular file: tail means per replicate.
-r report Report PDF.
-t Plot reports for all transcripts.
patsy-spike - estimate the number of sequencing errors in runs of bases.
usage: patsy-spike [-h] -n spike_sam -f ref [-w window] [-m max_errors_plot]
[-o out_pickle] [-q min_q] [-r report] [-l read_len]
[-pk pickle]
Estimate the number of sequencing errors in runs of bases (1.0).
optional arguments:
-h, --help show this help message and exit
-n spike_sam SAM file containing NVTR alignments.
-f ref Reference fasta with the spike-in sequences.
-w window Size of the flanking sequence around the run of As.
-m max_errors_plot Maximum number of errors for which to plot the length
distribution.
-o out_pickle Output pickle file.
-q min_q Mapping quality treshold (30).
-r report Report PDF.
-l read_len Read length.
-pk pickle Result pickle file.