5R software package
5R package – A software package accompanying a novel method of bacterial 16S rRNA profiling based on the 5R protocol. Five short regions along the 16S rRNA are amplified in multiplex and subsequently sequenced by an Illumina machine. Resulting reads from these regions are combined into a coherent solution using our Short MUltiple Regions Framework (SMURF) algorithm (Fuks et al. 2018).
Two options are provided: A Matlab code and a standalone executable version for Linux. The input to either methods is the same, and hence described below for the Matlab version followed by a short description of the standalone syntax.
Installation
Download the 5R package.
Usage
Open Matlab, change to the mFiles directory and run the following line:
main_5R(illumina_files_dir, db_dir, db_name, results_filename, kmer_len, generate_group_results)
where the inputs are:
illumina_files_dir – a (full) path to directory of fastq paired-end sequencing files of samples profiled by the 5R primers. The package supports two cases: (i) the fastq files corresponding to each sample are in a different sub-directory or (ii) fastq of different samples are mixed together in one folder. In the latter case the package will automatically split the files to folders per sample based on the the file name.
db_dir - the (full) path to the directory of data base kmers
db_name - the name of the data base (the default data base of 5R is name is "GreenGenes_201305")
results_filename - a full path to the results file name. Remark: Note that the directory in which results would be written must be created in advance.
kmer_len - the length of the k-mer to be applied for reconstruction. The maximal supported length is 160nt. To support higher resolution the kmer length should be set as high as possible, yet depending on read quality. Since read qualify often deteriorate with the reads' length, we suggest the length over which the typical read quality falls below 30. If not specified, the a default of 100nt is set.
generate_group_results - (this input parameter is optional) - this is a 0/1 indicator. When true, a folder named "groups" will be created inside the "resDir" folder for each sample and the full 16S sequences of all the reconstructed groups (i.e. amplicons) will be written to fasta files. A separate file will be created for each group containing all the sequences of the group.
Taxonomy
The taxonomy file named taxonomy_db.mat should be found in the db_dir along the data base file per region. If the file is not found, the reconstruction file will be created per group (instead of per species) with empty taxonomy.
Output
Two files are created:
results_filename.txt - A table of species' relative abundances (rows) in each sample (columns).
ReadCountStats_results_filename.txt - General statistics of the number reads that passed the quality filter, the number of reads matched to each regions etc.
Example
The directory "example_fastq" contains 5R sequencing files of two breast cancer FFPE samples. To perform SMURF reconstruction of these samples use the following command
main_5R('./example_fastq','./GG_5R', './example_results/5R_SMURF_example.txt',126);
Standalone vesion
The 5R package can also be applied on a linux machine without a Matlab license.
Download the following MCR file from the Mathworks site:
Following download double click on the installer and follow the instructions. This package is required for running standalone code compiled using Matlab.
For instructions regarding the MCR installation see:
https://www.mathworks.com/help/compiler/install-the-matlab-runtime.html
To run the 5R code use the following syntax
./5R_linux/run_main_5R.sh PATH_TO_RUNTIME illumina_files_dir db_dir db_name results_filename
where PATH_TO_RUNTIME is the directory in which the MCR was installed (e.g. /opt/apps/matlab/Runtime/v97). Replace PATH_TO_RUNTIME by your directory.
For example, the following syntax would work on the two above mentioned samples:
./5R_linux/run_main_5R.sh PATH_TO_RUNTIME ./example_fastq ./GG_5R GreenGenes_201305 ./example_results/5R_SMURF_example.txt 126
Contact us
For questions please email:
Garold Fuks: garoldf@gmail.com
Noam Shental: shental@openu.ac.il
Ravid Straussman: ravidst@weizmann.ac.il
References
Garold Fuks, Michael Elgart, Amnon Amir, Amit Zeisel, Peter J. Turnbaugh, Yoav Soen and Noam Shental. Combining 16S rRNA gene variable regions enables high-resolution microbial community profiling. Microbiome, 26;6(1):17, Jan 2018. doi: 10.1186/s40168-017-0396-x.