Version: 2.2.3
NOTE TEtranscripts and TEcount rely on specially curated GTF files, which are not packaged with this software due to their size. Please go to our website for instructions to download the curated GTF files.
TEtranscripts and TEcount takes RNA-seq (and similar data) and annotates reads to both genes & transposable elements. TEtranscripts then performs differential analysis using DESeq2.
Created by Ying Jin, Eric Paniagua, Oliver Tam & Molly Hammell, February 2014
Copyright (C) 2014-2021 Ying Jin, Eric Paniagua, Talitha Forcier, Oliver Tam & Molly Hammell
Contact: Oliver Tam ([email protected]) or Talitha Forcier ([email protected])
Python: 2.7.x or 3.2.x (tested on Python 2.7.11 and 3.7.7)
pysam: 0.9.x or greater
R: 2.15.x or greater
DESeq2: 1.10.x or greater
Download compressed tarball.
Unpack tarball.
Navigate into unpacked directory.
Run the following:
$ python setup.py install
If you want to install locally (e.g. /local/home/usr), run this command instead:
$ python setup.py install --prefix /local/home/usr
NOTE In the above example, you must add
/local/home/usr/bin
to the PATH variable, and
/local/home/usr/lib/pythonX.Y/site-packages
to the PYTHONPATH variable, where X refers to the major python version, and Y refers to the minor python version. (e.g. python2.7 if using python version 2.7.x, and python3.6 if using python version 3.6.x)
Many High Performance Compunting clusters (HPCs) have access to singularity which allows for the download and exectuation of containers, TEtranscripts also has a container through docker, it can be downloaded by singularity thusly:
singularity pull tetranscripts.sif docker://mhammelllab/tetranscripts:latest
Execution is then through singularity as well:
singularity exec tetranscripts.sif TEtranscripts -t <treatment sample> -c <control sample> --GTF <genic-GTF-file> --TE <TE-GTF-file>
usage: TEtranscripts -t treatment sample [treatment sample ...] -c control sample [control sample ...] --GTF genic-GTF-file --TE TE-GTF-file [optional arguments] Required arguments: -t | --treatment [treatment sample 1 treatment sample 2...] Sample files in group 1 (e.g. treatment/mutant), separated by space -c | --control [control sample 1 control sample 2 ...] Sample files in group 2 (e.g. control/wildtype), separated by space --GTF genic-GTF-file GTF file for gene annotations --TE TE-GTF-file GTF file for transposable element annotations Optional arguments: *Input/Output options* --format [input file format] Input file format: BAM or SAM. DEFAULT: BAM --stranded [option] Is this a stranded library? (no, forward, or reverse). no - Library is unstranded forward - "Second-strand" cDNA library (e.g. QIAseq stranded) reverse - "First-strand" cDNA library (e.g. Illumina TruSeq stranded) DEFAULT: no. --sortByPos Input file is sorted by chromosome position. --project [name] Prefix used for output files (e.g. project name) DEFAULT: TEtranscript_out --outdir [directory] Directory for output files. DEFAULT: current directory *Analysis options* --mode [TE counting mode] How to count TE: uniq (unique mappers only) multi (distribute among all alignments). DEFAULT: multi --minread [min_read] read count cutoff. DEFAULT: 1 -L | --fragmentLength [fragLength] Average length of fragment used for single-end sequencing DEFAULT: For paired-end, estimated from the input alignment file. For single-end, ignored by default. -i | --iteration maximum number of iterations used to optimize multi-reads assignment. DEFAULT: 100 -p | --padj [pvalue] FDR cutoff for significance. DEFAULT: 0.05 -f | --foldchange [foldchange] Fold-change ratio (absolute) cutoff for differential expression. DEFAULT: 1 *DESeq1 compatibility options* --DESeq Use DESeq (instead of DESeq2) for differential analysis. -n | --norm [normalization] Normalization method : DESeq_default (default normalization method of DESeq), TC (total annotated read counts), quant (quantile normalization). Only applicable if DESeq is used instead of DESeq2. DEFAULT: DESeq_default *Other options* -h | --help Show help message --verbose [number] Set verbose level. 0: only show critical messages 1: show additional warning messages 2: show process information 3: show debug messages DEFAULT: 2 --version Show program's version and exit
NOTE BAM files must be either unsorted or sorted by queryname. If the BAM files are sorted by position, please use the --sortByPos
option
If BAM files are unsorted, or sorted by queryname:
TEtranscripts --format BAM --mode multi -t RNAseq1.bam RNAseq2.bam -c CtlRNAseq1.bam CtlRNAseq.bam --project sample_nosort_test
If BAM files are sorted by coordinates/position:
TEtranscripts --sortByPos --format BAM --mode multi -t RNAseq1.bam RNAseq2.bam -c CtlRNAseq1.bam CtlRNAseq.bam --project sample_sorted_test
In our experience, we recommend around 20-30Gb of memory for analyzing human samples (hg19) with around 20-30 million mapped reads when running on a cluster.
usage: TEcount -b RNAseq BAM --GTF genic-GTF-file --TE TE-GTF-file [optional arguments] Required arguments: -b | --BAM alignment-file RNAseq alignment file (BAM preferred) --GTF genic-GTF-file GTF file for gene annotations --TE TE-GTF-file GTF file for transposable element annotations Optional arguments: *Input/Output options* --format [input file format] Input file format: BAM or SAM. DEFAULT: BAM --stranded [option] Is this a stranded library? (no, forward, or reverse). no - Library is unstranded forward - "Second-strand" cDNA library (e.g. QIAseq stranded) reverse - "First-strand" cDNA library (e.g. Illumina TruSeq stranded) DEFAULT: no. --sortByPos Input file is sorted by chromosome position. --project [name] Prefix used for output files (e.g. project name) DEFAULT: TEcount_out --outdir [directory] Directory for output files. DEFAULT: current directory *Analysis options* --mode [TE counting mode] How to count TE: uniq (unique mappers only) multi (distribute among all alignments). DEFAULT: multi -L | --fragmentLength [fragLength] Average length of fragment used for single-end sequencing DEFAULT: For paired-end, estimated from the input alignment file. For single-end, ignored by default. -i | --iteration maximum number of iterations used to optimize multi-reads assignment. DEFAULT: 100 *Other options* -h | --help Show help message --verbose [number] Set verbose level. 0: only show critical messages 1: show additional warning messages 2: show process information 3: show debug messages DEFAULT: 2 --version Show program's version and exit
NOTE BAM files must be either unsorted or sorted by queryname. If the BAM files are sorted by position, please use the --sortByPos
option
If BAM files are unsorted, or sorted by queryname:
TEcount --format BAM --mode multi -b RNAseq.bam --project sample_nosort_test
If BAM files are sorted by coordinates/position:
TEtranscripts --sortByPos --format BAM --mode multi -b RNAseq.bam --project sample_sorted_test
TEcount is better suited than TEtranscripts for usage in the cluster environment, as each sample (e.g. replicates of an experiment) can be quantified on separate nodes. The output can then be merged into a single count table for differential analysis. In our experience, we recommend around 20-30Gb of memory for analyzing human samples (hg19) with around 20-30 million mapped reads when running on a cluster.
TEtranscripts can perform transposable element quantification from alignment results (e.g. BAM files) generated from a variety of programs. Given the variety of experimental systems, we could not provide an optimal alignment strategy for every approach. Therefore, we recommend that users identify the optimal parameters for their particular genome and alignment program in order to get the best results.
When optimizing the alignment parameters, we recommend taking these points into consideration:
Allowing sufficient number of multi-mappers during alignment
Most alignment programs provide only 1 alignment per read by default. We recommend reporting multiple alignments per read. We have found that reporting a maximum of 100 alignments per read provides an optimal compromise between the size of the alignment file and recovery of multi-mappers in many genome builds. However, we highly suggest that users optimize this parameter for their particular experiment, as this could significantly improve the quality of transposable element quantification.
Optimizing alignment parameters for non-reference strains
It is common that the specific laboratory strains used in an experiment contains genomic variations not present in the reference strain. While this can be mitigated through allowing mismatches during alignments, certain lab strains (e.g. Drosophila melanogaster) have diverged significantly from the reference genomes. We highly recommend that users should refine their alignment procedures to better account for the expected variations between their lab strains and the reference genome, which will accordingly improve their analysis with TEtranscripts. Users can also align to a custom genome build specific to their organism, though they would need GTF annotations for genes and transposable elements that are compatible with their custom genome in order to utilize TEtranscripts. Please contact us if you require advice in generating these annotation files.
Paired end sequencing input
For paired-end libraries, it is recommended that only alignments from properly paired reads are present in the input BAM file. I.e., each read 1 alignment should only have a single read 2 alignment. For example, if read 1 matched 3 genomic locations (A, B, C), then if read 2 also match 3 genomic locations (A', B', C'), then all three pairs of alignments could be used (and should be in the BAM file). However, if alignment C of read 1 was matched with more than one alignment of read 2 (e.g. C' and C*), then alignment C should be discarded (as there are unmatched alignments between read 1 and read 2). STAR only outputs properly paired alignments by default, while Bowtie2 requires the --no-mixed
parameter to be used.
Specific recommendations when using STAR
STAR utilizes two parameters for optimal identification of multi-mappers --outFilterMultimapNmax and --outAnchorMultimapNmax. The author of STAR recommends that --winAnchorMultimapNmax should be set at twice the value used in --outFilterMultimapNmax, but no less than 50. In our study, we used the same number for both parameters (100), and found negligible differences in identifying multi-mappers. Upon further discussion with the author of STAR, we recommend that setting the same value for --winAnchorMultimapNmax and --outFilterMultimapNmax, though we highly suggest users test multiple values of --winAnchorMultimapNmax to identify the optimal value for their experiment.
TEtranscripts and TEcount are part of TEToolkit suite.
TEtranscripts is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with TEtranscripts. If not, see this website.
If using the software in a publication, please cite the following:
Jin Y, Tam OH, Paniagua E, Hammell M. (2015) TEtranscripts: a package for including transposable elements in differential expression analysis of RNA-seq datasets. Bioinformatics. 31(22):3593-9.