=======================================================================
Hypo is a package of both polisher, capable of correcting draft genomes, and an assembler to assemble a diploid genome. It exploits unique genomic kmers to both selectively polish only segments of contigs and to increase mapping accuracy.
Hypo is only available for Unix-like platforms (Linux and MAC OS). Currently, we provide the option of installation from source with CMake. CmakeLists is provided in the project root folder.
The following requirements are assumed to be installed (with path to their binaries available in $PATH).
- Zlib
- OpenMP
- GCC (>=7.3)
- KMC3
- minimap2
- samtools
- FlyE (optional, if draft assembly is not provided)
- pysam
- BioPython
First, you need to clone the repository recusrively (for the submodules)
git clone --recursive https://github.com/kensung-lab/hypo-assembler
After cloning, the easiest way to run the pipeline is to run build_all.sh. This will write a run_all directory, from where the run can be easily executed.
./build_all.sh
cd run_all
The following optional options are available for build_all.sh
Usage: ./build_all.sh <args>
** Optional parameters
-t <thread count>
The number of threads to run make
[Default] 10
-n
If present, will run optimized for native (i.e. -march=native)
-o <build directory>
Directory to put executables and scripts.
[Default] run_allUsage with run_all.sh is as follows:
Usage: ./run_all.sh <args>
** Mandatory arguments
-1 <short read file 1>
The file containing the first of the short read input
-2 <short read file 2>
The file containing the pair of the short read input
-l <long reads file>
The file containing long reads
** Optional parameters
-t <thread count>
The number of threads to run
[Default] 1
-o <output prefix>
The prefix for the output file names. The outputs will be <prefix>_1.fa and <prefix>_2.fa
[Default] hypo
-d <assembly draft>
The file containing the original draft. If not provided, FlyE will be run from the given long reads
[Default] None
-B <long reads mapping to draft>
Mappings of the long reads to the original draft as provided in -d. If not provided, minimap2 will be run to align the long reads
[Default] None
-s <estimated size>
Estimated size of the assembled genome. K/M/G suffix accepted
[Default] 3G
-T <tempdir>
Directory to write temporary files
[Default] temp
-m <sortmem>
The memory used for each samtools sorting thread. Samtools will use number of threads * sortmem memory.
[Default] 1G
-@ <sortthreads>
The threads used for each samtools sorting. Separate from the normal number of threads.
[Default] 10
-M <kmcmem>
The max memory for KMC, in GB
[Default] 12
-k <kmer length>
The length of solid kmers used in the method. Recommended is 17 for human genome.
[Default] 17We have included a small demo data to test the installation in demo/
For the puposes of this demo, the following external library is assumed:
Assuming build_all.sh is already run:
cd run_all
./run_all.sh -1 ../demo/il1.fq -2 ../demo/il2.fq -l ../demo/ont.fq.gz -t 40
This will run hypo-assembler on 40 threads.
The outputs will be hypo_1.fa and hypo_2.fa in the run_all directory.
To make sure the program runs properly, you can compare the output with the results in demo/expected_result/hypo_1.fa and demo/expected_result/hypo_2.fa, i.e.
diff hypo_1.fa ../demo/expected_result/hypo_1.fa
diff hypo_2.fa ../demo/expected_result/hypo_2.fa
If both commands execute without any output, then the program works properly.
We recommend running the full package of hypo-assembler with -d option to improve a quality of a draft genome. However, we also provide the option of running the polisher separate from the assembler.
To run them, simply run the binary "hypo" built from the polisher directory.
Usage: hypo <args>
** Mandatory args:
-r, --reads-short <str>
Input file name containing reads (in fasta/fastq format; can be compressed). A list of files containing file names in each line can be passed with @ prefix.
-d, --draft <str>
Input file name containing the draft contigs (in fasta/fastq format; can be compressed).
-b, --bam-sr <str>
Input file name containing the alignments of short reads against the draft (in bam/sam format; must have CIGAR information).
-c, --coverage-short <int>
Approximate mean coverage of the short reads.
-s, --size-ref <str>
Approximate size of the genome (a number; could be followed by units k/m/g; e.g. 10m, 2.3g).
** Optional args:
-B, --bam-lr <str>
Input file name containing the alignments of long reads against the draft (in bam/sam format; must have CIGAR information).
[Only Short reads polishing will be performed if this argument is not given]
-o, --output <str>
Output file name.
[Default] hypo_<draft_file_name>.fasta in the working directory.
-t, --threads <int>
Number of threads.
[Default] 1.
-p, --processing-size <int>
Number of contigs to be processed in one batch. Lower value means less memory usage but slower speed.
[Default] All the contigs in the draft.
-m, --match-sr <int>
Score for matching bases for short reads.
[Default] 5.
-x, --mismatch-sr <int>
Score for mismatching bases for short reads.
[Default] -4.
-g, --gap-sr <int>
Gap penalty for short reads (must be negative).
[Default] -8.
-M, --match-lr <int>
Score for matching bases for long reads.
[Default] 3.
-X, --mismatch-lr <int>
Score for mismatching bases for long reads.
[Default] -5.
-G, --gap-lr <int>
Gap penalty for long reads (must be negative).
[Default] -4.
-n, --ned-th <int>
Threshold for Normalised Edit Distance of long arms allowed in a window (in %). Higher number means more arms allowed which may slow down the execution.
[Default] 20.
-q, --qual-map-th <int>
Threshold for mapping quality of reads. The reads with mapping quality below this threshold will not be taken into consideration.
[Default] 2.
-i, --intermed
Store or use (if already exist) the intermediate files.
[Currently, only Solid kmers are stored as an intermediate file.].
-h, --help
Print the usage.
Other than raising issues in github, you can contact joshuac@comp.nus.edu.sg for specific issues.