chromsize
annoyed to have to create an index and cut it?
but first, how is this better than any other option? yeah, just check the image below.
googled 'get chromosome sizes from fasta', grab every command/tool I found and benchmarked it. surprisingly, you can lose 14 seconds of your life just waiting for those chrom sizes to be calculated. crazy.
> What's new on v.0.0.2?
> - now reads .gz!
> - CI implementation
## Usage
### Binary
``` rust
Usage: chromsize --fasta --output [-t ]
Arguments:
-f, --fasta : FASTA file
-o, --output : path to chrom.sizes
Options:
-t, --threads : number of threads [default: your max ncpus]
--help: print help
--version: print version
```
#### crate: [https://crates.io/crates/chromsize](https://crates.io/crates/chromsize)
### Installation
to install rust and use chromsize on your system follow this steps:
1. get installer: `curl https://sh.rustup.rs -sSf | sh` on unix, or go [here](https://www.rust-lang.org/tools/install) for other options
2. run `cargo install chromsize` (make sure `~/.cargo/bin` is in your `$PATH` before running it)
4. use `chromsize` with the required arguments
### Library
``` rust
use chromsize;
fn main() {
let input = PathBuf::new("/path/to/fasta.fa");
let output = PathBuf::new("/path/to/chrom.sizes");
let sizes: Vec<(String, u64)> = chromsize::chromsize(&input);
chromsize::write(sizes, &output)
}
```
### Python
build the port to install it as a pkg:
```
git clone https://github.com/alejandrogzi/chromsize.git && cd chromsize/py-chromsize
hatch shell
maturin develop --release
```
use it as a binary wrapper:
``` python3
import chromsize as cs
input = "/path/to/fasta.fa"
output = "/path/to/chrom.sizes"
cs.write_chromsizes(input, output)
```
or just get them directly
``` python3
import chromsize as cs
input = "/path/to/fasta.fa"
sizes = cs.get_chromsizes(input)
>>> print(sizes)
[
('chr1', 123),
('chr2', 456),
...
]
```
## Build
to build chromsize from this repo, do:
1. get rust
2. run `git clone https://github.com/alejandrogzi/chromsize.git && cd chromsize`
3. run `cargo run --release -- -i -o `
## Container image
to build the development container image:
1. run `git clone https://github.com/alejandrogzi/chromsize.git && cd chromsize`
2. initialize docker with `start docker` or `systemctl start docker`
3. build the image `docker image build --tag chromsize .`
4. run `docker run --rm -v "[dir_where_your_fa_is]:/dir" chromsize -f /dir/`
> ## Conda (not available yet)
> to use chromsize through Conda just:
> 1. `conda install chromsize -c bioconda` or `conda create -n chromsize -c bioconda chromsize`
> ## Nextflow (not available yet)
## Benchmark
do not believe me? run the benchmark on your own:
1. get .fa from any species you want (or download the ones I used from UCSC/NCBI)
2. install hyperfine: [https://github.com/sharkdp/hyperfine](https://github.com/sharkdp/hyperfine)
3. go to chromsize/bench and modify the `ASSEMBLIES` const with the .fa you've download
4. run `cargo run release --bin chromsize-benchmark -- -d /dir/where/my/fastas/are -a show-output ignore-failure`
here is all the info and metadata from my experiment:
### which tools I used?
| Tool | Command | Reference | Discussion |
|-----------|------------------------------------------------------------------------------------------------------------------|-----------|------------|
| seqkit | `seqkit fx2tab --length --name --header-line {assembly} > chrom.sizes` | [1] | [2] |
| chromsize | `target/release/chromsize -f {assembly} -o chrom.sizes` | [3] | |
| pyfaidx | `faidx {assembly} -i chromsizes > chrom.sizes` | [4] | [5] |
| samtools | `samtools faidx {assembly} && wait \| cut -f1,2 {assembly}.fai > chrom.sizes` | [6] | [5] |
| faSize | `faSize -detailed -tab {assembly} > chrom.sizes` | [7] | |
| awk1 | `awk '/^>/ {if (seqlen){print seqlen}; print ;seqlen=0;next; } { seqlen += length($0)}END{print seqlen}' {assembly} > chrom.sizes` | [8] | [9] |
| awk2 | `awk '/^>/{if (l!=") print l; print; l=0; next}{l+=length($0)}END{print l}' {assembly} > chrom.sizes` | [8] | [9] |
| bioawk1 | `bioawk -c fastx '{print > $name ORS length($seq)}' {assembly} > chrom.sizes` | [10] | [9] |
| awk3 | `cat {assembly} \| awk '$0 ~ > {if (NR > 1) {print c;} c=0;printf substr($0,2,100) "\t"; } $0 !~ ">" {c+=length($0);} END { print c; }' > chrom.sizes` | [8] | [11] |
| bioawk2 | `bioawk -c fastx '{ print $name, length($seq) }' < {assembly} > chrom.sizes` | [10] | [2] |
[1]: https://github.com/shenwei356/seqkit
[2]: https://www.biostars.org/p/118954/
[3]: https://github.com/alejandrogzi/chromsize
[4]: https://github.com/mdshw5/pyfaidx
[5]: https://www.biostars.org/p/173963/
[6]: https://github.com/samtools/samtools
[7]: https://github.com/ENCODE-DCC/kentUtils/blob/master/bin/linux.x86_64/faSize
[8]: https://www.gnu.org/software/gawk/manual/gawk.html
[9]: https://stackoverflow.com/questions/23992646/sequence-length-of-fasta-file
[10]: https://github.com/lh3/bioawk
[11]: https://www.danielecook.com/generate-fasta-sequence-lengths/
### detailed data?
| Species | Assembly | Size (Gb) | chromsize | seqKit | awk1 | awk2 | awk3 | bioawk1 | bioawk2 | faSize | pyfaidx | samtools |
|----------------|-------------|-----------|-----------|---------------|---------------|---------------|---------------|----------------|----------------|---------------|----------------|----------------|
| S. cerevisiae | R64 | 0.01 | 0.004 | 0.016 (X 4.0) | 0.043 (X 10.7)| 0.043 (X 10.7)| 0.05 (X 12.5) | 0.03 (X 7.5) | 0.03 (X 7.5) | 0.054 (X 13.5)| 0.101 (X 25.2) | 0.064 (X 16.0) |
| C. elegans | ce11 | 0.10 | 0.02 | 0.103 (X 5.1) | 0.409 (X 20.4)| 0.408 (X 20.4)| 0.492 (X 24.6)| 0.274 (X 13.7) | 0.274 (X 13.7) | 0.426 (X 21.3)| 0.225 (X 11.2) | 0.472 (X 23.6) |
| D. melanogaster| dm6 | 0.14 | 0.028 | 0.147 (X 5.2) | 0.581 (X 20.7)| 0.583 (X 20.8)| 0.714 (X 25.5)| 0.426 (X 15.2) | 0.418 (X 14.9) | 0.633 (X 22.6)| 0.337 (X 12.0) | 0.667 (X 23.8) |
| D. rerio | danRer11 | 1.37 | 0.22 | 0.742 (X 3.4) | 6.815 (X 31.0)| 6.803 (X 30.9)| 8.216 (X 37.3)| 3.946 (X 17.9) | 3.95 (X 18.0) | 7.202 (X 32.7)| 3.029 (X 13.8) | 7.633 (X 34.7) |
| C. familiaris | canFam4 | 2.48 | 0.311 | 1.209 (X 3.9) | 10.158 (X 32.7)| 10.124 (X 32.6)| 12.206 (X 39.2)| 6.55 (X 21.1) | 6.518 (X 21.0) | 10.671 (X 34.3)| 4.741 (X 15.2) | 11.394 (X 36.6)|
| H. sapiens | GRCh38 | 3.10 | 0.43 | 1.696 (X 3.9) | 12.393 (X 28.8)| 12.432 (X 28.9)| 13.681 (X 31.8)| 7.414 (X 17.2) | 7.284 (X 16.9) | 13.102 (X 30.5)| 6.37 (X 14.8) | 14.074 (X 32.7)|
| B. bombina | aBomBom1 | 9.80 | 1.554 | 8.501 (X 5.5) | 41.676 (X 26.8)| 41.696 (X 26.8)| 49.064 (X 31.6)| 24.202 (X 15.6)| 24.374 (X 15.7)| 43.856 (X 28.2)| 19.755 (X 12.7)| 45.387 (X 29.2)|
| A. mexicanum | AmbMex60DD | 28.20 | 3.327 | 14.375 (X 4.3)| 118.923 (X 35.7)| 118.422 (X 35.6)| 137.781 (X 41.4)| 57.626 (X 17.3)| 57.591 (X 17.3)| 121.257 (X 36.4)| 54.82 (X 16.5) | 128.374 (X 38.6)|
| P. annectens | PAN1.0 | 40.10 | 4.606 | 18.664 (X 4.1)| 167.85 (X 36.4)| 165.701 (X 36.0)| 196.833 (X 42.7)| 91.747 (X 19.9)| 91.924 (X 20.0)| 170.475 (X 37.0)| 77.707 (X 16.9)| 181.562 (X 39.4)|
### how well performs with .gz?
#### CHM13-T2T.fa.gz
| Tool | Cores | Time |
|-----------|----------------------------|------------------------|
| seqkit | 16 | 18.993 s ± 0.132 s |
| chromsize | default (max_cpus: 16) | 7.631 s ± 0.010 s |
| seqkit | default (4) | 18.525 s ± 0.520 s |
| chromsize | 4 | 8.035 s ± 0.077 s |
| seqkit | 2 | 18.535 s ± 0.376 s |
| chromsize | 2 | 8.284 s ± 0.030 s |
