Skip to content

Commit

Permalink
Merge pull request DaehwanKimLab#304 from DaehwanKimLab/3ndoc
Browse files Browse the repository at this point in the history 
update hisat-3n webpage
  • Loading branch information
@imzhangyun
imzhangyun authored Jul 6, 2021
2 parents c3cf368 + 9d5260e commit 3304fff
Showing 1 changed file with 54 additions and 37 deletions.
91 changes: 54 additions & 37 deletions docs/_pages/hisat-3n.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -46,59 +46,76 @@ A few notes:
Install
------------

git clone https://github.com/DaehwanKimLab/hisat2.git
cd hisat2
git clone https://github.com/DaehwanKimLab/hisat2.git hisat-3n
cd hisat-3n
git checkout -b hisat-3n origin/hisat-3n
make


Make sure that you are in the `hisat-3n` branch


Build a 3N index with `hisat-3n-build`
Build a HISAT-3N index with `hisat-3n-build`
-----------
`hisat-3n-build` builds a 3N-index, which contains two hisat2 indexes, from a set of DNA sequences. For standard 3N-index,
each index contains 16 files with suffix `.3n.*.*.ht2`.
For repeat 3N-index, there are 16 more files in addition to the standard 3N-index, and they have the suffix
`.3n.*.rep.*.ht2`.
These files constitute the hisat-3n index and no other file is needed to alignment reads to the reference.

* Example for standard HISAT-3N index building:
`hisat-3n-build genome.fa genome`
* `--base-change <chr1,chr2>` argument is required for `hisat-3n-build` and `hisat-3n`.
Provide which base is converted in the sequencing process to another base. Please enter
2 letters separated by ',' for this argument. The first letter(chr1) should be the converted base, the second letter(chr2) should be
the converted to base. For example, during slam-seq, some 'T' is converted to 'C',
please enter `--base-change T,C`. During bisulfite-seq, some 'C' is converted to 'T', please enter `--base-change C,T`.
* Different conversion types may build the same hisat-3n index. Please check the table below for more detail.
Once you build the hisat-3n index with C to T conversion (for example BS-seq).
You can align the T to C conversion reads (for example SLAM-seq reads) with the same index.

* Example for repeat HISAT-3N index building (require 256 GB memory):
`hisat-3n-build --repeat-index genome.fa genome`

It is optional to make the graph index and add SNP or spicing site information to the index, to increase the alignment accuracy.
for more detail, please check the [HISAT2 manual].
| Conversion Types | HISAT-3N index suffix |
|:----------------------------------:|:-----------------------------:|
|C -> T<br>T -> C<br>A -> G<br>G -> A|.3n.CT.\*.ht2 <br>.3n.GA.\*.ht2|
|A -> C<br>C -> A<br>G -> T<br>T -> C|.3n.AC.\*.ht2 <br>.3n.TG.\*.ht2|
|A -> T<br>T -> A |.3n.AT.\*.ht2 <br>.3n.TA.\*.ht2|
|C -> G<br>G -> C |.3n.CG.\*.ht2 <br>.3n.GC.\*.ht2|

#### Examples:
# Build the standard HISAT-3N index (with C to T conversion):
hisat-3n-build --base-change C,T genome.fa genome

# Build the repeat HISAT-3N index (with T to C conversion, require 256 GB memory for human genome index):
hisat-3n-build --base-change T,C --repeat-index genome.fa genome

It is optional to make the graph index and add SNP or spice site information to the index, to increase the alignment accuracy.
The graph index building may require more memory than the linear index building.
For more detail, please check the [HISAT2 manual].

[HISAT2 manual]:https://daehwankimlab.github.io/hisat2/manual/

# Standard HISAT-3N integrated index with SNP information
hisat-3n-build --exons genome.exon genome.fa genome
#### Examples:
# Build the standard HISAT-3N index integrated index with SNP information
hisat-3n-build --snp genome.snp genome.fa genome

# Standard HISAT-3N integrated index with splicing site information
# Build the standard HISAT-3N integrated index with splice site information
hisat-3n-build --ss genome.ss genome.fa genome

# Repeat HISAT-3N integrated index with SNP information
hisat-3n-build --repeat-index --exons genome.exon genome.fa genome
# Build the repeat HISAT-3N index integrated index with SNP information
hisat-3n-build --repeat-index --snp genome.snp genome.fa genome

# Repeat HISAT-3N integrated index with splicing site information
# Build the repeat HISAT-3N integrated index with splice site information
hisat-3n-build --repeat-index --ss genome.ss genome.fa genome

Alignment with `hisat-3n`
------------
After we build the HISAT-3N index, you are ready to use `hisat-3n` for alignment.
HISAT-3N uses the HISAT2 argument but has some extra arguments. Please check [HISAT2 manual] for more detail.

For human genome reference, HISAT-3N requires about 9GB for alignment with standard 3N-index and 10.5 GB for repeat 3N-index.

* `--base-change <chr1,chr2>`
Provide which base is converted in the sequencing process to another base. Please enter
2 letters separated by ',' for this argument. The first letter(chr1) should be the converted base, the second letter(chr2) should be
the converted to base. For example, during slam-seq, some 'T' is converted to 'C',
please enter `--base-change T,C`. During bisulfite-seq, some 'C' is converted to 'T', please enter `--base-change C,T`.
If you want to align non-converted reads to the regular HISAT2 index, do not use this option.

* `--index/-x <hisat-3n-idx>`
The index for HISAT-3N. The basename is the name of the index files up to but not including the suffix `.3n.*.*.ht2` / etc.
Expand All @@ -114,14 +131,14 @@ For human genome reference, HISAT-3N requires about 9GB for alignment with stand
`--un-conc`, and `--al-conc`. We will fix this problem in future version. Please do not use `--unique-only` if you want to use the hisat2 options above.

#### Examples:
* Single-end slam-seq reads (T to C conversion, RNA) alignment with standard 3N-index:
`hisat-3n --index genome -f -U read.fa -S alignment_result.sam --base-change T,C --no-repeat-index`

* Paired-end bisulfite-seq reads (C to T conversion, DNA) alignment with repeat 3N-index:
`hisat-3n --index genome -f -1 read_1.fa -2 read_2.fa -S alignment_result.sam --base-change C,T --repeat --no-spliced-alignment`

* Single-end TAPS reads (have C to T conversion， RNA) alignment with repeat 3N-index:
`hisat-3n --index genome -q -U read.fq -S alignment_result.sam --base-change C,T --repeat`
# Single-end slam-seq reads (T to C conversion, RNA) alignment with the standard 3N-index:
hisat-3n --index genome -f -U read.fa -S alignment_result.sam --base-change T,C --no-repeat-index
# Paired-end bisulfite-seq reads (C to T conversion, DNA) alignment with the repeat 3N-index:
hisat-3n --index genome -f -1 read_1.fa -2 read_2.fa -S alignment_result.sam --base-change C,T --repeat --no-spliced-alignment
# Single-end TAPS reads (have C to T conversion， RNA) alignment with the repeat 3N-index:
hisat-3n --index genome -q -U read.fq -S alignment_result.sam --base-change C,T --repeat



Expand Down Expand Up @@ -179,17 +196,17 @@ Generate 3N-conversion-table with `hisat-3n-table`:


#### Examples:
* Generate 3N conversion table for bisulfite sequencing data:
`hisat-3n-table -p 16 --alignments sorted_alignment_result.sam --ref genome.fa --output-name output.tsv --base-change C,T`

* Generate 3N-conversion-table for TAPS data and only count base in CpG island and uniquely aligned:
`hisat-3n-table -p 16 --alignments sorted_alignment_result.sam --ref genome.fa --output-name output.tsv --base-change C,T --CG-only --unique-only`

* Generate 3N conversion table for bisulfite sequencing data from sorted BAM file:
`samtools view -h sorted_alignment_result.bam | hisat-3n-table --ref genome.fa --alignments - --output-name output.tsv --base-change C,T`

* Generate 3N conversion table for bisulfite sequencing data from unsorted BAM file:
`samtools sort alignment_result.bam -O sam | hisat-3n-table --ref genome.fa --alignments - --output-name output.tsv --base-change C,T`
# Generate the 3N-conversion-table for bisulfite sequencing data:
hisat-3n-table -p 16 --alignments sorted_alignment_result.sam --ref genome.fa --output-name output.tsv --base-change C,T
# Generate the 3N-conversion-table for TAPS data and only count base in CpG island and uniquely aligned:
hisat-3n-table -p 16 --alignments sorted_alignment_result.sam --ref genome.fa --output-name output.tsv --base-change C,T --CG-only --unique-only
# Generate the 3N-conversion-table for bisulfite sequencing data from sorted BAM file:
samtools view -h sorted_alignment_result.bam | hisat-3n-table --ref genome.fa --alignments - --output-name output.tsv --base-change C,T
# Generate the 3N-conversion-table for bisulfite sequencing data from unsorted BAM file:
samtools sort alignment_result.bam -O sam | hisat-3n-table --ref genome.fa --alignments - --output-name output.tsv --base-change C,T


#### Note:
Expand Down

0 comments on commit 3304fff

Please sign in to comment.