mike

Requirements

C++ version: C++11

R Language version: 3.2.2 or above but not exceed 4.0

python version: 3.6 or higher

kmc(https://github.com/refresh-bio/KMC.git)

Attention: if you use C++17, it maybe reports errors.

Installation

If you show error message when installing R package, you can skip the installation of R package and build the tree manually in the last step.

git clone https://github.com/Argonum-Clever2/mike.git
cd src
make
# install R package
Rscript install.r

Tutorial

The Fisrt Step

You need to install KMC in advance, and add kmc to PATH. Then, run command below. The file will be processed into a kmer file. Or you can input the kmer file directly, just skip the step.

# help
python kmc.py --help
# run
python kmc.py -f file1 file2 file3 file4 file5 file6 -d dirpath

If the python kmc.py script gives an error, you can also just run the kmc command to process all kmc-acceptable file formats into kmer files.

# kmc--first step
## single-end sequencing file
kmc -k21 -t10 INPUT.fastq OUTPUT_PREFIX DIRPATH
## paired-end sequencing file--write two sequencing files to a file list(INPUT.fastq.list)
kmc -k21 -t10 @INPUT.fastq.list OUTPUT_PREFIX DIRPATH

# kmc-second step
kmc_tools transform OUTPUT_PREFIX sort . dump -s OUTPUT_PREFIX.txt

kmer file

the format of a kmer file should like below. Each line consists of a 21-mer string and a number representing the frequency of occurrence of that 21-mer string, separated by a '\t'.

AAAAAAAAAAAAAAAAAAAAA 255

AAAAAAAAAAAAAAAAAAAAC 255

AAAAAAAAAAAAAAAAAAAAG 255

AAAAAAAAAAAAAAAAAAAAT 255

AAAAAAAAAAAAAAAAAAACA 255

AAAAAAAAAAAAAAAAAAACC 255

AAAAAAAAAAAAAAAAAAACG 255

... ...

filelist

The filelist means the file that includes a list of kmer files. The filelist needs to include the absolute path and filename.

ABSOLUTE_PATH/kmer_name_file_1

ABSOLUTE_PATH/kmer_name_file_2

ABSOLUTE_PATH/kmer_name_file_3

... ...

The Second Step

SKETCH

The second step is to process the kmer files in the filelist as sketched files, note that you need to enter the absolute paths

./mike sketch -t 10 -l ABSOLUTE_PATH/filelist -d DIRPATH

sketched filelist

The sketched file is the file obtained in the previous step, which ends with 'jac'. The sketched filelist is the file that includes a list of sketched file.

ABSOLUTE_PATH/sketched_file_1.jac

ABSOLUTE_PATH/sketched_file_2.jac

ABSOLUTE_PATH/sketched_file_3.jac

... ...

The Third Step

the Jaccard coefficient

compute the jaccard coefficient for pairwire, and then will generate the file named jaccard.txt in destination_path

./mike compute -l ABSOLUTE_PATH/sketched_filelist -L ABSOLUTE_PATH/sketched_filelist -d DIRPATH

the evolutionary distance

compute the evolutionary distance，and then will generate the file named dist.txt in destination_path

./mike dist -l ABSOLUTE_PATH/sketched_filelist -L ABSOLUTE_PATH/sketched_filelist -d DIRPATH

The Final Step

construction the phylogenetic tree

using the evulutionary distance (dist.txt) to construct the phylogenetic tree without branch length. the file titled dist.txt was generated from the evolutionary distance

Rscript draw.r -f dist.txt -o dist.nwk

If the final step encounters an error, you can manually construct the phylogenetic tree by opening RStudio, downloading the ape package, and inputting the file dist.txt.

install.packages("ape")
library(ape)
tree <- read.csv("absolute_path/dist.txt", sep='\t', header = TRUE, row.names = 1)
treedist <- as.dist(tree)
# bionj
tree <- bionj(treedist)
# nj
tree <- nj(treedist)
# output
write.tree(tree, "tree.nwk")