dependency

• Csh: required by MUMmer
• Blasr: please install this and make sure you can execute the command blasr -h
• Python 2.7

The following dependencies are included in this repo for easier installation:

• MUMmer (with bug fixed)
• Lemon
• mhap
• muscle

Install

Run python install.py, this will install smsc to the build/local/bin directory inside this repository. If you want to install it to <local>/bin/, run python install.py --prefix <local>

Known Issue

• gcc-9.1, or the clang linked to this version of gcc has a problem compiling Lemon (see Issue #2). If this is the case, please either use a lower version gcc (gcc-4.8, gcc-6, gcc-7, gcc-8.3 have been tested), or install Lemon by yourself. If you choose to install Lemon by yourself, please use the command ls $(dirname $(which dimacs-solver))/../share/lemon/cmake/LEMONConfig.cmake to make sure the LEMONConfig.cmake file exists.

run

Go to the directory of smsc (default: build/local/bin)

Type ./smsc -h to see the running options

output

• <prefix>.final.bin: the scaffolding in binary format, including the counting for each nucleotide A simplified parser will be added to this repository
• <prefix>.final.fasta: the scaffolding in fasta format
• <prefix>.dis: disassembling locations
  • Each line is a triple delimited by a space draft_id start_loc end_loc
  • draft_id: the index of the draft assembly (start from 0)
  • start_loc: the start location of the corresponding draft assembly (start from 1)
  • end_loc: the end location of the corresponding draft assembly (right included)
• <prefix>.path: reassembling path
  • There are multiple blocks in this file. Each block is a path
  • In each block, the first line is tuple delimited by a space path_id number_of_verticess
  • In the following number_of_vertices lines, each line is a quadruple draft_id start_loc end_loc direction
  • The first three entry are identical to that in <prefix>.dis
  • direction: if 0, then it's we use the forward direction. If 1, we use the reverse complement.
  • Notice: start_loc and end_loc always denote the original locations in the draft assembly. So if direction=1, then we should extract the corresponding subsequence in the draft assembly and then compute the reverse complement.

sample data

The following is a sample of a simulated data.

• The E. Coli K12 (ground truth) can be downloaded here
• The manually mutated genome is in sample/E_coli_k12.mut.fasta
• The error-corrected Nanopore long reads can be downloaded here

A simple way to run of the program would be

smsc -p <prefix> <draft_assembly> <long_reads>

The program can also run in multithreading mode my setting -j <# of threads>.

The default alignment tool is Nucmer. If -E is set, the alignment tool would be Blasr.

For example, suppose that the sample data E_coli_k12.mut.fasta and the Nanopore file Nanopore.fasta are in the current directory, the command

smsc -j 8 -p Nanopore E_coli_k12.mut.fasta Nanopore.fasta

will use 8 threads to produce a corrected-assembly. The result is in the file Nanopore.final.fasta. There is also a binary file Nanopore.final.bin containing the counting of each nucleotide in each scaffolds of final output.

---

Update 2.2.1

• N.B.: We restrict the running instances of MHAP to at most 8 on purpose. The mhap.jar keeps crashing if there are 16 instances.

Update 2.1.0

• deal with N's in the genome sequence

Update 2.0.1

• new path cover algorithm
• Usage: as before (change path in run_nucmer.sh)

---