Strobealign is a single or paired-end short-read aligner using syncmer-thinned strobemers. Strobealign is multithreaded and implements both alignment (SAM) and mapping (PAF). It is very fast, see experiments for single-end and paired-end alignment in the preprint. Strobealign is not recommended for very short reads (roughly <=80nt).
You can acquire precompiled binaries for Linux and Mac OSx from the release page.
If you want to compile from the source, you need to have a newer g++ and zlib installed. Then do the following:
git clone https://github.com/ksahlin/StrobeAlign
cd StrobeAlign
# Needs a newer g++ version. Tested with version 8 and upwards.
g++ -std=c++14 main.cpp source/index.cpp source/xxhash.c source/ksw2_extz2_sse.c -lz -fopenmp -o StrobeAlign -O3 -mavx2
If you have zlib installed, and the zlib.h file is in folder /path/to/zlib/include and the libz.so file in /path/to/zlib/lib but you get
main.cpp:12:10: fatal error: zlib.h: No such file or directory
#include <zlib.h>
^~~~~~~~
compilation terminated.
add -I/path/to/zlib/include -L/path/to/zlib/lib to the compilation, that is
g++ -std=c++14 -I/path/to/zlib/include -L/path/to/zlib/lib main.cpp source/index.cpp source/xxhash.c source/ksw2_extz2_sse.c -lz -fopenmp -o StrobeAlign -O3 -mavx2
Strobealign v0.1 and up comes with a parameter -r read_length [100, 150, 200, 250, 300] that sets suitable parameters for the given rough read length estimate. Specifically, it sets parameters -k, -l and -u. Valid values of -r currently are 100, 150, 200, 250, and 300. If not specified, it defaults to 150. The value of r does not have to match the exact read length.
For alignment to SAM file:
StrobeAlign -r <read_length> -o <output.sam> ref.fa reads.fa
For mapping to PAF file (option -x):
StrobeAlign -r <read_length> -x -o <output.sam> ref.fa reads.fa
- Add option to separate build index and perform alignment in separate steps.
Kristoffer Sahlin. Faster short-read mapping with strobemer seeds in syncmer space. bioRxiv, 2021. doi:10.1101/2021.06.18.449070. Preprint available here.
Major update to algorithm. See release page.
- Takes care of negative alignment coordinate bug.
- Minimal value for repetitive seed filtering implemented. Previoulsy, the top fraction of
-f (0.0002)seeds were filtered regardless of how repetitive they were. Assume-ffiltered everything aboveXoccurences. The new version filters seeds with occurences overmax(X, 30). This threshold is usually not active for human asX>40for human.
- Bugfix. Takes care of segmentation fault bug in paired-end mapping mode (-x) when none of the reads have NAMs.
- Implements a paired-end alignment mode.
- Implements a rescue mode both in SE and PE alignment modes (described in preprint v2).
- Changed to symmetrical strobemer hashvalues due to inversions (described in preprint v2).
- Implements multi-threading.
- Allow reads in fast[a/q] format and gzipped files through kseqpp library.
The aligner used for the experiments presented in the preprint (v1) on bioRxiv. Only single threaded alignment and aligns reads as single reads (no PE mapping).
GPL v3.0, see LICENSE.txt.