apa is a Python framework for processing and analysing 3'-end targeted sequence data to study alternative polyadenylation. apa depends on pybio (basic handling of annotated genomes), RNAmotifs2 (analysis of regulatory motif clusters) and other open-source software (DEXSeq, STAR short-read aligner).
The inclusive nature of the framework, together with novel integrative solutions (differential polyA site usage and RNA-protein binding via RNA-maps, cluster motif analysis), results in the following computational capabilities:
Since this python package has several dependencies (pybio for genome download and manupulation, for example), the easiest way to try out the package with an example dataset is by running the docker image (provided with the Dockerfile).
Clone this repository (git clone https://github.com/grexor/apa.git) and run build.sh to build the Docker image (you need to have Docker installed).
This will build a Docker image with apa, pybio and all other dependencies installed. It will also create a "data" folder on your local drive (inside the same folder where build.sh is) where larger genome and result files will be stored. You can change the location of the data folder by modifying the build.sh script.
To login to the system (user apauser), simply run the run_apauser.sh script. You are now running the Docker container with all required dependencies and software to run the apa example provided.
To directly start the example processing, follow brief instructions in Running the example.
We provide an example Lexogen Quantseq Reverse sequencing run consisting of 6 experiments (3 HEK293, 3 TDP-43 KD) from the publication:
In total 6 FASTQ files. The reads will be downloaded and processed in the Docker container by the docker/example.sh script. Reads for the example were preselected to match only the ones mapping to the chr22 of the hg38 assembly.
Data is organized inside libraries. Each library contains several experiments. Each experiment is represented by a FASTQ file. The sequencing (library) data is stored in the apa.path.data_folder. The location of the data folder can be changed by editing apa/config/config.txt and adding a line with apa.path.data_folder="/path/to/data_folder".
# structure the data folder on apa
->data_folder [folder]
->example [folder] # folder of sequencing library with id example
->annotation.tab [file] # annotation TAB separated file (see below for structure)
->example.config # config file for the entire library (see below for structure)
->e1 [folder] # experiment 1
->example_e1.fastq.gz [file] # FASTQ file of experiment 1
->m1 [folder] # several mappings can be performed (diverse parameters), named m1, m2, ...
->e2 [folder] # experiment 2
->example_e2.fastq.gz [file] # FASTQ file of experiment 2
->example2 [folder] # folder of sequencing library with id example2
->example3 [folder] # folder of sequencing library with id example3
...The annotation.tab file is present for each library (always with the same name) and annotates the experiments within the library. It's a TAB separated file with the following structure:
# first line is always a comment
exp_id species map_to method condition replicate
1 hg38chr22 hg38chr22 lexrev HEK293 1
2 hg38chr22 hg38chr22 lexrev HEK293 2
3 hg38chr22 hg38chr22 lexrev HEK293 3
4 hg38chr22 hg38chr22 lexrev KD 1
5 hg38chr22 hg38chr22 lexrev KD 2
6 hg38chr22 hg38chr22 lexrev KD 3Each line represents an experiment.
exp_id = experiment id (integer number)
species = species, descriptive (human, mouse, ms, hg, etc.)
map_to = id of genome (linked to pybio package) for read mapping
method = id of method
condition = descriptive
replicate = descriptiveThe library config file is named with libraryid.config. In our example, since our library has id example, the file is example.config. The minimum structure of the library config file is as follows:
name:HEK-239
notes:Library with 3'-end Lexogen Quantrev sequencing of HEK-293 cell lines
method:lexrev
genome:hg38chr22
map_to:hg38chr22
seq_type:single
tags:
status:
columns:[['Condition', 'condition'], ['Replicate', 'replicate']]
columns_display:[['Condition', 'condition'], ['Replicate', 'replicate']]Using DEXSeq, we can identify APA genes by comparing sets of control vs. test experiments. The comparisons are stored in the apa.path.comps_folder, each comparison is defined by the config file.
Structure of the comparison data folder is:
->comps_folder [folder]
->example [folder] # folder of the comparison with id example
->example.config [file] # config file of the comparison
->example.pairs_de.tab [file] # TAB file with results, one line per gene, regulated and control pairs of polyA sitesWe run the comparative analysis with apa.comps -comps_id example (for the example provided).
TAB separated file:
id experiments name
c1 example_e1 HEK293_1
c2 example_e2 HEK293_2
c3 example_e3 HEK293_3
t1 example_e4 KD_1
t2 example_e5 KD_2
t3 example_e6 KD_3
control_name:HEK293
test_name:KD
site_selection:DEX
polya_db:example
poly_type:["strong"]
# at least half of the experiments at a specific polyA site need to count at least 5 reads
presence_thr:2.0
cDNA_thr:5
analysis_type:apa
method:lexrev
genome:hg38chr22In the example above, 3 control and 3 test experiments, all coming from library example.
To run the provided example, start ~/apa/docker/example.sh inside the Docker container. This will download the chr22 of the hg38 genome assembly, download and map the 6 example experiments (3 HEK293 and 3 TDP-43 KD) to the hg38 genome (only chromosome 22). It will build a polyA database from the aligned reads, estimate read counts at the identified polyA sites and also provide gene expression.
Content of docker/example.sh:
# map all library experiments with STAR
apa.map.lib -lib_id example
# create BED files from mapped reads according to protocol
apa.bed.multi -lib_id example
# make configuration file for polyA database (includes all experiments in the library)
apa.polya.makeconfig -lib_id example
# create polyA database
apa.polya -poly_id example
# compute gene expression using htseq-count
apa.bed.gene_expression -lib_id example
# compute polyA site counts according to provided mapping and polyA database
apa.bed.multi -lib_id example -type expression -poly_id example -upstream 10 -downstream 10
# compute comparative analysis
apa.comps -comps_id exampleSince the setup and analysis of data with this package can be time consuming, we developed an interactive online platform expressRNA. You can upload your experiment FASTQ files, annotate and perform analysis on the expressRNA server if desired.
The best way to install apa, pybio and other dependencies on your own server is simply to follow the Dockerfile. In case of problems, open an Issue on this repository page.
For a full list of dependencies, please refer to the Dockerfile.
BiocManager::install("edgeR")BiocManager::install("DEXSeq")RUN pip3 install HTSeqpip3 install regexapa is developed and supported by Gregor Rot in collaboration with several research laboratories worldwide.
The development started in 2009 when Tomaž Curk and Gregor Rot wrote the first prototype of apa. In 2013, Gregor Rot refactored and further developed the code, also establishing expressRNA, a web application for exploring results of alternative polyadenylation analysis.
High-resolution RNA maps suggest common principles of splicing and polyadenylation regulation by TDP-43
Rot, G., Wang, Z., Huppertz, I., Modic, M., Lenče, T., Hallegger, M., Haberman, N., Curk, T., von Mering, C., Ule, J.
Cell Reports , Volume 19 , Issue 5 , 1056 - 1067
Use the issues page to report issues and leave suggestions.