LoRMA
Contact: leena.salmela@cs.helsinki.fi

--------
Overview
--------

LoRMA is a tool for correcting sequencing errors in long reads.

---------
Reference
---------

L. Salmela, R. Walve, E. Rivals, and E. Ukkonen: Accurate
selfcorrection of errors in long reads using de Bruijn graphs.
(To appear in Bioinformatics)

-------------------
System Requirements
-------------------

LoRMA has been tested on systems running Linux on a X86_64
architecture. LoRMA uses GATB library
(http://gatb-core.gforge.inria.fr/index.html) for the de Bruijn graph
implementation. The library is included in the LoRMA package.
Compiling LoRMA requires gcc version 4.5 or newer and cmake.

------------
Installation
------------

Unpack the LoRMA package.
For compiling LoRMA run

    mkdir build;  cd build;  cmake ..;  make

The binary called LoRMA can then be found in the build directory.

LoRMA is designed to be used with LoRDEC. To install LoRDEC, see
further details at

http://atgc.lirmm.fr/lordec/

For best results the strict mode of LoRDEC should be used. This mode
is available starting from LoRDEC version 0.6.

Edit the variable LORDECDIR in lorma.sh to point to your LoRDEC
installation directory.

-----
Usage
-----

The script lorma.sh runs LoRDEC iteratively, trims and splits the
resulting initial corrected reads, and finally runs LoRMA on these
trimmed and split reads.

Usage: lorma.sh [parameters] <long read fasta files>

[-s] [-n] [-start <19> -end <61> -step <21> -threads <6> -friends <7> -k <19>] *.fasta

Parameters:
-s                   save sequence data of the intermediate LoRDEC steps
-n                   do not run LoRDEC steps
-start <int>         k for the first LoRDEC step (default: 19)
-end <int>           upper limit for k for the LoRDEC steps (default: 61)
-step <int>          the increase of k between the LoRDEC steps (default: 21)
-k <int>             k value for running LoRMA (default: 19)
-friends <int>       the number of reads in multiple alignments in LoRMA (default: 7)
-threads <int>       the number of threads used by LoRMA (default: 6)

Note that LoRMA uses a substantial amount of memory per thread.
Therefore be careful when increasing the number of threads used by
LoRMA.

-------------------
Parameters of LoRMA
-------------------

Note that normally the script lorma.sh should be used to run LoRMA.
The LoRMA executable is run as follows:

LoRMA [parameters]

Required parameters:
-reads <file>         file(s) of long reads for correction
-output <file>        output file for corrected reads
-discarded <file>     output file for discarded reads

Optional parameters:
-k <int>              kmer length (default: 31)
-nb-cores <int>       number of threads to use (default: 1)
-friends <int>        number of friends (default: 7)
-bestfriends <int>    number of best friends (default: 3)
-verbose <int>        verbosity level (default: 1)

-------
Example
-------

lorma.sh reads.fasta

------------------
New in Version 0.4
------------------

Some bug fixes:
- compatibility with GCC 5.4
- fixed a bug in lorma.sh

------------------
New in Version 0.3
------------------

First public version.