Release News
Fix a bug related to annotation file processing. This is a rare case when a gene occurs at multiple chromosomes but closed enough in the gtf file. Related to an issue #42
- docker pull ruolinliu/strawberry:v1.1.1
- Adding ref_gene_id and ref_gene_name fields to output gtf file, if reference guided assembly is used. These are taken from the matched reference loci.
- When running with
--no-quantflag, TPM and FPKM parts from the output gtf file will be omitted. - The
--output-diroption is changed to--output-gtfoption. - The logfile now by default is saved to
/tmp/strawberry.log - Fix mem bugs. Some details see: #33 and #36.
- Update -g option. When using reference annotations, the closest gene id/symbol will be added to the output gtf file. If no reference gene is found, a new gene id will be assigned. Credit goes to Chris for suggesting this new feature #30.
- Bug fix for loading transcripts.
- Fix a bug when outputing fragment context using -f.
- Allow the fragment context output to contain FPKM column which is passed to Strawberry2.
- Fix a bug when loading the last transcript is out of bound. This is an edge case when used the
-goption. Credit goes to 47Lies for reporting the bug and providing the test case.
- Support Pacbio CCS reads.
- Update logging.
- Fix bug for 1bp intron from gff file.
- Fix a bug when a bam file contains number only contig/scaffold names. This bug caused dropping all reads.
- Add Strand-specific options
--fr assume stranded library fr-secondstrand
--rf assume stranded library rf-firststrand
Figure from One Tipe Per Day
- Fix a bug when the reference sequence names (e.g., contig or chromosome names) are long. This bug affects assembled_transcripts.gtf and sometimes make the first column unreadable.
- Add TPM (Transcript Per Million mapped reads) to the output.
Strawberry is a C++ program for fast and accurate ab initio transcript reconstruction and quantification from RNA-seq data. It is written in C++11 and is available as open source software. Strawberry leverages the speed and accuracy of transcript assembly and quantification in such a way that processing 10 million simulated reads (after alignment) requires only 2 minutes using a single thread while achieving over 92% correlation with the ground truth, making it the state-of-the-art method.
Strawberry is a genome-guided transcript-level assembler and quantification tool. It takes aligned RNA-Seq data in BAM format and outputs a gene annotation files in gff format with estimated transcripts abundances. Using alignment file as input allows Strawberry to take advantages of the latest reference genome (if possible, a finished and high-quality one) and stat-of-the-art splice-awareness aligners. The application of a fast flow network algorithm, for assembly speeds up the construction of transcript-based models. The resulting reduced data representation improves quantification of the different isoforms. Strawberry is also able to account for various sequencing bias that is intrinsic to the RNA-seq experiment. For paired-end RNA-Seq, Strawberry empirically infers the insert length distribution from the reads that are mapped uniquely and concordantly. If half-mapped reads exist, Strawberry generates the other ends based on the mapped orientations and the insert length distribution.
Strawberry consists of two modules: assembly module and quantification module. The two modules work in a sequential manner. The assembly module parses aligned reads into splicing graphs, and it uses network flow algorithms to selected the most likely transcripts. The assembly module can take an existing gene annotation file as a guide and this practice usually leads to better assembly result. The quantification module uses a statistical model, more accurately a Latent Class Model, to assign ambiguous reads to transcripts. Strawberry simultaneous estimates the transcript abundances and corrects for sequencing bias through the EM steps. The quantification model can be executed along without doing an assembly. This requires a user to provide an additional gene annotation file in gff3 or gtf format.
Strawberry is now published in PLoS Computational Biology.
https://github.com/ruolin/strawberry/releases/download/1.0.2/strawberry
-
A C++14 conformant compiler (currently tested with GCC>=5.3).
- Here is wiki about how to install GCC, https://gcc.gnu.org/wiki/InstallingGCC.
- And for ubuntu user, please refer to the post for how to install gcc 5+, https://askubuntu.com/questions/618474/how-to-install-the-latest-gcurrently-5-1-in-ubuntucurrently-14-04.
-
Strawberry uses CMake(3.5+) build system to check and install dependencies and to compile and install Strawberry.
-
Samtools. Strawberry uses an older version of Samtools(v0.1.19) as a dependent library.
-
If a different version of Samtools (currently Samtools split into 3 projects now and has been updated to v1.3) has been installed, or your system does not have Samtools installed at all, please let Strawberry automatically download and install Samtools for you. You DONOT need to do anything.
-
If you have installed Samtools v0.1.19 and want to use it, please make sure that libbam.a is in the Samtools source directory and set an environmental variable SAMTOOLS_ROOT to that location.
- If you use csh or tcsh, at the shell prompt, enter
setenv SAMTOOLS_ROOT /path/to/samtools-source-dir.- If you use bash, at the shell promet, enter
export SAMTOOLS_ROOT /path/to/samtools-source-dir.
-
-
Some user reported that libncurses5 is missing in the system for Ubuntu 16.04LTS and 18.04LTS
- sudo apt-get install libncurses5-dev
git clone https://github.com/ruolin/strawberry.git
Note: Precompiled binary is available on the [release page] (https://github.com/ruolin/Strawberry/releases) for the latest version.
sh cmake.shcd buildmake
WARNING: using parallel build (-j option) might cause errors.
The executable file will be found in the Strawberry/bin directory.
You can add this directory to your PATH variable to complete the installation by
export PATH=/path/to/your/intallation/bin/:$PATH
docker pull ruolinliu/strawberry:v1.1.1
after pulling the docker image, run, for example:
docker run -v /home/ruolin/:/mnt ruolinliu/strawberry:v1.1.1 /home/strawberry/bin/strawberry /mnt/git/strawberry/examples/geuvadis_300/sample_01.sorted.bam -o /mnt/sample_01.gtf
Running Strawberry is relatively easy. You need to have an alignment file in BAM format. This step can be done using any splice-awareness aligner, e.g. Tophat 2, GSNAP, HISAT 2. The BAM file needs to be sorted according to the genomic positions. If you use Tophat, the default output is already sorted. For other software, you might have to sort their outputs before running Strawberry. This can be done using Samtools command samtools sort.
I provide a toy bam file for testing the instalation.
Under the root directory of your installation, type bin/strawberry examples/geuvadis_300/sample_01.sorted.bam to run the program on the default parameters. This bam files contains read from 300 genes. And if everything is fine, Strawberry will finish in seconds.
Strawberry can be run in three different modes. The default mode is to do assembly and quantification without reference annotation.
bin/strawberry examples/geuvadis_300/sample_01.sorted.bam -o output.gtf -p 8
You can also run strawberry with the help of reference annotation. In this mode, Strawberry uses the known gene models to guide assembly. Novel isoform discovery are still allowed. If you have a good quality reference annotation but also want to detect novel isoforms, this is the ideal mode.
bin/strawberry examples/geuvadis_300/sample_01.sorted.bam -o output.gtf -g reference.gtf -p 8
If you truth your gene annotation completely and do not want to assembly any new isoforms, you can skip the assembly step and just do quantification against provided gene models.
bin/strawberry examples/geuvadis_300/sample_01.sorted.bam -o output.gtf -g reference.gtf -r -p 8
Good luck!
For the choice of parameters and their meanings type strawberry without any argument for help information.