liuchuwei / canano

caNano

Dependence

Genome

Table of Contents

1. Intallation
2. Usage
3. References
4. License
5. Contact

Intallation

1.Clone the project

git clone https://github.com/liuchuwei/caNano.git

1.Install conda environment

conda env create -f caNano.yaml

2.prepare tookit:

check and modify the tookit.py file (in 'utils' directory).

Usage

1.Basecalling

python 01.basecalling.py -i $fast5 -o $out

2.Resguiggle

preprocess

conda activate tombo
python 02.resquiggle_pre.py -f $fast5 -o $out

annotate_raw_with_fastqs

cat *.fastq > merge.fastq
python 03.resquiggle.py preprocess annotate_raw_with_fastqs \
--fast5-basedir $single \
--fastq-filenames $merge_fastq \
--overwrite \
--processes 8

resquiggling

python 03.resquiggle.py resquiggle $fast5 $reference \
--rna \
--corrected-group RawGenomeCorrected_000 \
--basecall-group Basecall_1D_000 \
--overwrite \
--processes 16 \
--fit-global-scale \
--include-event-stdev

3.Minimap

python 04.minimap.py -i <directory of fastq files> -o <output directory> -r <path of reference>

4.Eventalign: preprocess data for m6anet model, you can skip this step if you don't use m6anet model.

python 05.eventalign.py -f <directory of fast5 files> -o <output directory> \
 -fq <path of fastq> -r <path of reference> -bam <path of bam files> -o <output directory>

5.m6a detection

(1) caNano

activate environment

conda activate caNano

preprocess

python caNano.py preprocess --single $single_fast5 -o $output -g $genome.fa -r $transcript.fa -i $gene2transcripts.txt -b $bam

train_read

python caNano.py train_read --mod %mod.tsv --unmod %unmod.tsv --out %output

train_site

python caNano.py predict_read --input %feature.tsv --output %output --model %model

(2) Tombo

activate environment

conda activate tombo

predict

python tombo.py detect_modifications de_novo --fast5-basedirs <directory of fast5 files> \
--statistics-file-basename <output name> \
--corrected-group RawGenomeCorrected_000 \
--processes 16

output

python tombo.py text_output browser_files --fast5-basedirs <directory of fast5 files> \
--statistics-filename <output name>  \
--browser-file-basename wt_rrach \
--genome-fasta <path of reference> \
--motif-descriptions RRACH:3:m6A \
--file-types coverage dampened_fraction fraction \
--corrected-group RawGenomeCorrected_000

(3) Mines

activate environment

conda activate Mines

tidy tombo result

awk '{if($0!=null){print $0}}' wt.fraction_modified_reads.plus.wig > wt.wig
wig2bed < wt.wig > wt.fraction_modified_reads.plus.wig.bed --multisplit=mines

predict

python Mines.py --fraction_modified $tombo/wt.fraction_modified_reads.plus.wig.bed \
--coverage $tombo/wt.coverage.plus.bedgraph \
--output wt.bed \
--ref $ref \
--kmer_models $MINES/Final_Models/names.txt

(4) m6anet

data preprocess

m6anet dataprep --eventalign wt_eventalign.txt
--out_dir wt
--n_processes 16
--readcount_max 2000000

predict

m6anet inference --input_dir wt
--out_dir run/wt
--n_processes 16

(5) Nanom6A

list all fast5 files

find single -name "*.fast5" >files.txt

extracting signals

extract_raw_and_feature_fast --cpu=20 --fl=files.txt -o result --clip=10

predict

predict_sites --cpu 20 -i result -o result_final -r data/cc_ref.fa -g data/cc_ref.fa -b data/gene2transcripts.txt --model Nanom6A/model

(6) Eligos

python ELGOS rna_mod -i <bam file> -reg <bed files> -ref <REFERENCE> -m <rBEM5+2 model> -p <output file prefix> \
-o <output file directory> --sub_bam_dir <SUB_BAM_DIR> \
--max_depth 2000000 --min_depth 5 --esb 0 --oddR 1 --pval 1 -t 16

(7) Epinano

extract features

python 07.Epinano_Variants.py -R $ref -b <wt / ko bam file> -s <path to sam2tsv> -n 16 -T t

slide features

python 07.Epinano_slide_feature.py <per.site.csv> 5

preict

python 07.Epinano_Predict.py \
--model $Epinano/models/rrach.q3.mis3.del3.linear.dump \
--predict wt.plus_strand.per.site.5mer.csv \
--columns 8,13,23 \
--out_prefix wt

License

Distributed under the GPL-2.0 License License. See LICENSE for more information.

Contact

[email protected]

Reference