SPUMONI is a software tool for performing rapid read classifications on sequencing reads using a read's matching statistics (or a related quantity called pseudo-matching lengths).
SPUMONI is based on another software tool called MONI which is a MEM-finder and aligner for pan-genomes. MONI uses prefix-free parsing of the text [2][3] to build the Burrows-Wheeler Transform (BWT) of the reference collection, the suffix array (SA) samples at the beginning and end of each run, and the threshold positions[1].
For more details on installation/using SPUMONI, please refer to the wiki page which covers those areas in detail.
There are two ways to use SPUMONI. The easier way would be to use the Docker/Singularity images which have SPUMONI pre-installed and ready to use. The other way would be to clone this GitHub repo as shown below and build from the source files.
(1) Use Docker or Singularity Containers
docker pull oma219/spumoni:latest
docker run oma219/spumoni spumoni -h
Since docker requires root permissions, singularity can be used instead on shared clusters as shown below.
singularity pull docker://oma219/spumoni:latest
singularity run spumoni_latest.sif spumoni -h
Both sets of commands above will download the containers and run spumoni to print out the sub-commands that are available to run.
(2) Build from Source
To build from source, it requires the following dependences: zlib, cmake, gcc. See the wiki for detailed information.
git clone https://github.com/oma219/spumoni
cd spumoni
mkdir build && cd build
cmake ..
make install
export SPUMONI_BUILD_DIR=$(pwd)
# use spumoni ...After running that last command above, in the build/ directory there will be a spumoni executable to use.
After installing SPUMONI on your machine, the first step would be to build an index over the reference you want to use for your experiment. This reference will be a FASTA file (and it can be a multi-FASTA for pan-genomes). SPUMONI allows you to either build it over a single FASTA file, or you can specify a list of genomes that you want to include in the index. See the wiki for more details.
For example, if you would want to do host depletion, you could index a human genome and build both the matching statistic and pseudo-matching lengths index using the command below:
./spumoni build -r human_genome.fa -M -P -m -o /path/to/indexThe command above will build an index for the reference file you provide. This command uses -M and -P which means it will build an index for computing matching statistics (MSs), and another index for computing pseudo-matching statistics (PMLs). Our experiments show that using PMLs are more accurate at binary classification while also being ~3x faster, therefore that would be our recommendation. The -o flag provides an output prefix for SPUMONI to use so it can build the index and any intermediate files it needs.
Once you have an index for desired experiment, you can use the spumoni run command to generate either MSs or PMLs for each read against the reference file that you just indexed. Now at this step, you provide the output prefix that was used to build the index.
./spumoni run -r /path/to/index -p reads.fa -P -cThis command uses -P for computing PMLs (if you want MSs, use -M instead) which will be used to classify the reads. Additionally, the command uses the -c option to write out the classifications to a report file.
If you run into any issues or have any questions, please feel free to reach out to us either (1) through GitHub Issues or (2) reach out to me at omaryfekry [at] gmail.com
MONI is the Finnish word for multi.
SPUMONI stands for Streaming PseUdo MONI.
If you use the SPUMONI or MONI tools in your research project, please cite:
Ahmed, O., Rossi, M., Kovaka, S., Schatz, M. C., Gagie, T., Boucher, C., & Langmead, B. (2021). Pan-genomic Matching Statistics for Targeted Nanopore Sequencing. iScience, 102696.
Rossi, M., Oliva, M., Langmead, B., Gagie, T., & Boucher, C. (2021). MONI: A Pangenomics Index for Finding MEMs. Proc. RECOMB.
[1] Hideo Bannai, Travis Gagie, and Tomohiro I, "Refining ther-index", Theoretical Computer Science, 812 (2020), pp. 96–108
[2] Christina Boucher, Travis Gagie, Alan Kuhnle and Giovanni Manzini, "Prefix-Free Parsing for Building Big BWTs", In Proc. of the 18th International Workshop on Algorithms in Bioinformatics (WABI 2018).
[3] Christina Boucher, Travis Gagie, Alan Kuhnle, Ben Langmead, Giovanni Manzini, and Taher Mun. "Prefix-free parsing for building big BWTs.", Algorithms for Molecular Biology 14, no. 1 (2019): 13.