Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 4 additions & 5 deletions Cargo.toml
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
[package]
name = "ms2rescore-rs"
version = "0.4.0-1"
version = "0.4.1"
edition = "2021"

# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
Expand All @@ -10,10 +10,9 @@ crate-type = ["cdylib"]

[features]
default = ["thermo"]

thermo = ["mzdata/thermo"]

[dependencies]
pyo3 = "0.20.0"
mzdata = "0.33.0"
timsrust = "0.3.0"
pyo3 = { version = "0.23.3", features = ["anyhow"] }
mzdata = "0.39.0"
timsrust = "0.4.1"
21 changes: 11 additions & 10 deletions src/lib.rs
Original file line number Diff line number Diff line change
@@ -1,17 +1,17 @@
mod file_types;
mod ms2_spectrum;
mod parse_mzdata;
mod parse_timsrust;
mod precursor;
mod ms2_spectrum;

use std::collections::HashMap;

use pyo3::exceptions::{PyException, PyValueError};
use pyo3::prelude::*;

use file_types::{match_file_type, SpectrumFileType};
use precursor::Precursor;
use ms2_spectrum::MS2Spectrum;
use precursor::Precursor;

/// Check if spectrum path matches a supported file type.
#[pyfunction]
Expand All @@ -27,9 +27,10 @@ pub fn get_precursor_info(spectrum_path: String) -> PyResult<HashMap<String, Pre
let file_type = match_file_type(&spectrum_path);

let precursors = match file_type {
SpectrumFileType::MascotGenericFormat | SpectrumFileType::MzML | SpectrumFileType::MzMLb | SpectrumFileType:: ThermoRaw => {
parse_mzdata::parse_precursor_info(&spectrum_path)
}
SpectrumFileType::MascotGenericFormat
| SpectrumFileType::MzML
| SpectrumFileType::MzMLb
| SpectrumFileType::ThermoRaw => parse_mzdata::parse_precursor_info(&spectrum_path),
SpectrumFileType::BrukerRaw => parse_timsrust::parse_precursor_info(&spectrum_path),
SpectrumFileType::Unknown => return Err(PyValueError::new_err("Unsupported file type")),
};
Expand All @@ -46,9 +47,10 @@ pub fn get_ms2_spectra(spectrum_path: String) -> PyResult<Vec<ms2_spectrum::MS2S
let file_type = match_file_type(&spectrum_path);

let spectra = match file_type {
SpectrumFileType::MascotGenericFormat | SpectrumFileType::MzML | SpectrumFileType::MzMLb | SpectrumFileType:: ThermoRaw => {
parse_mzdata::read_ms2_spectra(&spectrum_path)
}
SpectrumFileType::MascotGenericFormat
| SpectrumFileType::MzML
| SpectrumFileType::MzMLb
| SpectrumFileType::ThermoRaw => parse_mzdata::read_ms2_spectra(&spectrum_path),
SpectrumFileType::BrukerRaw => parse_timsrust::read_ms2_spectra(&spectrum_path),
SpectrumFileType::Unknown => return Err(PyValueError::new_err("Unsupported file type")),
};
Expand All @@ -59,10 +61,9 @@ pub fn get_ms2_spectra(spectrum_path: String) -> PyResult<Vec<ms2_spectrum::MS2S
}
}


/// A Python module implemented in Rust.
#[pymodule]
fn ms2rescore_rs(_py: Python, m: &PyModule) -> PyResult<()> {
fn ms2rescore_rs(m: &Bound<'_, PyModule>) -> PyResult<()> {
m.add_class::<Precursor>()?;
m.add_class::<MS2Spectrum>()?;
m.add_function(wrap_pyfunction!(is_supported_file_type, m)?)?;
Expand Down
37 changes: 19 additions & 18 deletions src/parse_mzdata.rs
Original file line number Diff line number Diff line change
@@ -1,7 +1,6 @@
use std::collections::HashMap;

use mzdata::params::ParamValue;
use mzdata::mz_read;
use mzdata::{params::ParamValue, prelude::*, MZReader};

use crate::ms2_spectrum::MS2Spectrum;
use crate::precursor::Precursor;
Expand Down Expand Up @@ -49,25 +48,27 @@ impl From<mzdata::spectrum::MultiLayerSpectrum> for MS2Spectrum {
pub fn parse_precursor_info(
spectrum_path: &str,
) -> Result<HashMap<String, Precursor>, std::io::Error> {
mz_read!(spectrum_path.as_ref(), reader => {
reader.filter(|spectrum| spectrum.description.ms_level == 2)
.filter_map(|spectrum| {
spectrum.description.precursor.as_ref()?;
Some((spectrum.description.id.clone(), Precursor::from(&spectrum)))
})
.collect::<HashMap<String, Precursor>>()
})
let reader = MZReader::open_path(spectrum_path)?;
Ok(reader
.filter(|spectrum| spectrum.description.ms_level == 2)
.filter_map(|spectrum| {
spectrum.description.precursor.as_ref()?;
Some((spectrum.description.id.clone(), Precursor::from(&spectrum)))
})
.collect::<HashMap<String, Precursor>>())
}

/// Read MS2 spectra from spectrum files with mzdata
pub fn read_ms2_spectra(
spectrum_path: &str,
) -> Result<Vec<MS2Spectrum>, std::io::Error> {
mz_read!(spectrum_path.as_ref(), reader => {
reader.filter(|spectrum| spectrum.description.ms_level == 2)
.map(MS2Spectrum::from)
.collect::<Vec<MS2Spectrum>>()
})
pub fn read_ms2_spectra(spectrum_path: &str) -> Result<Vec<MS2Spectrum>, std::io::Error> {
let mut reader = MZReader::open_path(spectrum_path)?;
if let MZReader::ThermoRaw(inner) = &mut reader {
inner.set_centroiding(true);
}

Ok(reader
.filter(|spectrum| spectrum.description.ms_level == 2)
.map(MS2Spectrum::from)
.collect::<Vec<MS2Spectrum>>())
}

fn get_charge_from_spectrum(spectrum: &mzdata::spectrum::MultiLayerSpectrum) -> Option<usize> {
Expand Down
51 changes: 26 additions & 25 deletions src/parse_timsrust.rs
Original file line number Diff line number Diff line change
Expand Up @@ -3,8 +3,8 @@ use std::collections::HashMap;
use crate::ms2_spectrum::MS2Spectrum;
use crate::precursor::Precursor;

impl From<timsrust::ms_data::Precursor> for Precursor {
fn from(precursor: timsrust::ms_data::Precursor) -> Self {
impl From<timsrust::Precursor> for Precursor {
fn from(precursor: timsrust::Precursor) -> Self {
Precursor {
mz: precursor.mz,
rt: precursor.rt,
Expand All @@ -15,8 +15,8 @@ impl From<timsrust::ms_data::Precursor> for Precursor {
}
}

impl From<timsrust::ms_data::Spectrum> for MS2Spectrum {
fn from(spectrum: timsrust::ms_data::Spectrum) -> Self {
impl From<timsrust::Spectrum> for MS2Spectrum {
fn from(spectrum: timsrust::Spectrum) -> Self {
MS2Spectrum::new(
spectrum.index.to_string(),
spectrum.mz_values.iter().map(|mz| *mz as f32).collect(),
Expand All @@ -25,7 +25,7 @@ impl From<timsrust::ms_data::Spectrum> for MS2Spectrum {
.iter()
.map(|intensity| *intensity as f32)
.collect(),
Some(Precursor::from(spectrum.precursor)),
spectrum.precursor.map(Precursor::from),
)
}
}
Expand All @@ -34,35 +34,36 @@ impl From<timsrust::ms_data::Spectrum> for MS2Spectrum {
pub fn parse_precursor_info(
spectrum_path: &str,
) -> Result<HashMap<String, Precursor>, std::io::Error> {
let reader = timsrust::FileReader::new(spectrum_path)
let reader = timsrust::readers::SpectrumReader::new(spectrum_path)
.map_err(|e| std::io::Error::new(std::io::ErrorKind::Other, e.to_string()))?;

Ok(reader
.read_all_spectra()
let spectra = reader
.get_all()
.into_iter()
.filter(|spectrum| {
matches!(
spectrum.precursor,
timsrust::ms_data::Precursor { .. }
)
})
.map(|spectrum| {
(
spectrum.index.to_string(),
Precursor::from(spectrum.precursor),
)
.collect::<Result<Vec<_>, _>>()
.map_err(|e| std::io::Error::new(std::io::ErrorKind::Other, e.to_string()))?;

let precursor_info = spectra
.into_iter()
.filter_map(|spectrum| match spectrum.precursor {
Some(precursor) => Some((spectrum.index.to_string(), Precursor::from(precursor))),
None => None,
})
.collect::<HashMap<String, Precursor>>())
.collect::<HashMap<_, _>>();

Ok(precursor_info)
}

/// Read MS2 spectra from spectrum files with timsrust
pub fn read_ms2_spectra(spectrum_path: &str) -> Result<Vec<MS2Spectrum>, std::io::Error> {
let reader = timsrust::FileReader::new(spectrum_path)
let reader = timsrust::readers::SpectrumReader::new(spectrum_path)
.map_err(|e| std::io::Error::new(std::io::ErrorKind::Other, e.to_string()))?;

Ok(reader
.read_all_spectra()
let spectra = reader
.get_all()
.into_iter()
.map(MS2Spectrum::from)
.collect())
.collect::<Result<Vec<_>, _>>()
.map_err(|e| std::io::Error::new(std::io::ErrorKind::Other, e.to_string()))?;

Ok(spectra.into_iter().map(MS2Spectrum::from).collect())
}
Loading