SEMQuant

SEMQuant is a set of proteomics quantification pipelines developed by the Biocomputing Research Group. The repository bundles the tools used by the workflows — Raxport, Sipros Ensemble, Percolator, Philosopher, IonQuant and DIA-NN — together with the helper scripts that glue them together.

Two pipelines are provided:

Pipeline	Directory	Acquisition	Quantification
SEMQuant-DDA	`SEMQuant-DDA/`	Thermo DDA	IonQuant (MS1 LFQ + MBR)
SEMQuant-Astral	`SEMQuant-Astral/`	Orbitrap Astral	DIA-NN (spectral library)

Repository layout

SEMQuant-DDA/
├── Raxport                     # Thermo .raw → FT2 converter (.NET / mono)
├── Sipros_OpenMP               # database search engine
├── SiprosEnsembleOMP           # Sipros Ensemble search engine
├── configs/                    # SiprosConfig*.cfg templates
├── Scripts/                    # Python/shell post-processing (filtering, assembly)
│   └── runSiprosFiltering.sh   # wrapper: tabulating → filtering → assembly → clustering
├── SE2IQ/                      # Sipros Ensemble → IonQuant/ProteinProphet converters
├── IonQuant-1.10.12.jar        # quantification (with jfreechart / batmass-io jars)
├── bruker/  thermo/            # vendor libraries required by IonQuant
├── yeast_ups.fasta             # example database (yeast + UPS1)
└── runSEMQuant.sh              # end-to-end example driver

SEMQuant-Astral/
├── percolator_3_6_4/           # PSM rescoring (linux/ and windows/ builds)
├── philosopher_v5.1.0_linux_amd64/   # protein inference + FDR control
├── linux_diann.zip             # DIA-NN 1.8.2 beta 8 (Git LFS; unzip before use)
├── script3/                    # Python 3 port of the Sipros post-processing scripts
└── SE_IQ_Astral/               # converters: Percolator/Sipros → ProteinProphet, IonQuant

Requirements

Create the conda environments used throughout the workflows:

conda create -n py2  -c conda-forge scikit-learn python=2.7
conda create -n py3  -c conda-forge python=3.10
conda create -n mono -c conda-forge mono

Raxport and ThermoRawFileParser require .NET / mono.
SEMQuant-DDA/Scripts/ targets Python 2.7 (py2); SEMQuant-Astral/script3/ is the Python 3 port (py3).
IonQuant requires a Java 11+ runtime.
Some post-processing steps use R.
linux_diann.zip is stored with Git LFS — run git lfs install && git lfs pull after cloning, then unzip it before running the DIA-NN step.

Tools not bundled in this repository and which must be obtained separately: ThermoRawFileParser, and the FragPipe speclib/gen_con_spec_lib.py spectral-library generator with its unimod.xml.

Conventions

Every command below assumes these two variables are exported:

export work_dir=/PATH_TO_YOUR_WORK_DIR      # where raw data and results live
export tool_dir=/PATH_TO/SEMQuant           # this repository

Individual pipeline sections then use $tool_dir/SEMQuant-DDA or $tool_dir/SEMQuant-Astral.

SEMQuant-DDA

DDA workflow: raw conversion → Sipros Ensemble database search → PSM filtering and protein assembly → IonQuant label-free quantification with match-between-runs.

1. Set up the working directory

export work_dir=/PATH_TO_YOUR_WORK_DIR
export dda_dir=$tool_dir/SEMQuant-DDA

mkdir -p $work_dir/{raw,samples,results}

2. Download example raw files

cd $work_dir/raw
# yeast-UPS1 2 fmol replicates (PXD002099)
wget ftp://ftp.pride.ebi.ac.uk/pride/data/archive/2020/01/PXD002099/*_2fmol*.raw

This yields three sample raw files.

3. Convert raw to FT2

conda activate mono
$dda_dir/Raxport -i $work_dir/raw -o $work_dir/raw

4. Generate the target–decoy database

conda activate py2
python $dda_dir/Scripts/sipros_prepare_protein_database.py \
    -i $dda_dir/yeast_ups.fasta \
    -o $work_dir/yeast_ups_rev.fasta \
    -c $dda_dir/configs/SiprosConfig_yeast.cfg

Replace the input/output FASTA with your own database, then point the configuration file at the new decoy-appended database:

FASTA_Database = /PATH_TO_YOUR_WORK_DIR/yeast_ups_rev.fasta

5. Database search

SiprosEnsembleOMP searches either a single MS2 file (-f) or every MS2 file in a directory (-w). Use -g instead of -c to sweep a directory of configuration files, and SiprosEnsembleOMP -h for the full option list.

export OMP_NUM_THREADS=24

# single MS2 file
$dda_dir/SiprosEnsembleOMP -o $work_dir/results \
    -f $work_dir/raw/YOUR_MS2.FT2 \
    -c $dda_dir/configs/SiprosConfig_yeast.cfg

# all MS2 files in a directory
$dda_dir/SiprosEnsembleOMP -o $work_dir/results \
    -w $work_dir/raw \
    -c $dda_dir/configs/SiprosConfig_yeast.cfg

Search results are written to the output directory as .Spe2Pep.txt files.

6. Split results into per-sample folders

cd $work_dir
for sample in results/*Spe2Pep.txt; do
    name=$(basename "${sample%%.*}")
    mkdir -p samples/"$name"
    cp "$sample" samples/"$name"/
done

7. Filter PSMs and assemble proteins

runSiprosFiltering.sh chains PSM tabulating, ensemble filtering, protein assembly and SIP-mode clustering:

conda activate py2
for folder in $work_dir/samples/*/; do
    $dda_dir/Scripts/runSiprosFiltering.sh \
        -in "$folder" \
        -c  $dda_dir/configs/SiprosConfig_yeast.cfg \
        -o  "$folder"
done

The individual steps can also be run directly:

for folder in $work_dir/samples/*/; do
    python $dda_dir/Scripts/sipros_ensemble_filtering.py \
        -i "$folder" -c $dda_dir/configs/SiprosConfig_yeast.cfg -o "$folder"
    python $dda_dir/Scripts/sipros_peptides_assembling.py \
        -w "$folder" -c $dda_dir/configs/SiprosConfig_yeast.cfg
done

8. Convert results for IonQuant

conda activate py3
for folder in $work_dir/samples/*/; do
    name=$(basename "$folder")
    python $dda_dir/SE2IQ/SE2prophet.py   "$folder/$name.psm.txt" "$folder" "$name"
    python $dda_dir/SE2IQ/SE2Ionquant.py  $work_dir/yeast_ups_rev.fasta "$folder"
done

# build the IonQuant file list, then copy the modification-mass list next to it
python $dda_dir/SE2IQ/generate_filelist.py
cp $dda_dir/modmasses_ionquant.txt $work_dir/samples/

filelist_ionquant.txt and modmasses_ionquant.txt in SEMQuant-DDA/ are templates — edit --specdir in the file list to point at your raw/ directory.

9. Quantify with IonQuant

java -Xmx21G \
  -Dlibs.bruker.dir=$dda_dir/bruker \
  -Dlibs.thermo.dir=$dda_dir/thermo \
  -cp "$dda_dir/jfreechart-1.5.3.jar:$dda_dir/batmass-io-1.30.0.jar:$dda_dir/IonQuant-1.10.12.jar" \
  ionquant.IonQuant \
  --threads 23 --perform-ms1quant 1 --perform-isoquant 0 \
  --isotol 20.0 --isolevel 2 --isotype tmt10 --ionmobility 0 \
  --site-reports 1 --minexps 1 --mbr 1 --maxlfq 1 --requantify 1 \
  --mztol 10 --imtol 0.05 --rttol 0.4 \
  --mbrmincorr 0 --mbrrttol 1 --mbrimtol 0.05 --mbrtoprun 40 \
  --ionfdr 0.01 --proteinfdr 0.01 --peptidefdr 0.01 \
  --normalization 1 --minisotopes 2 --minscans 3 --minions 2 --minfreq 0 \
  --writeindex 0 --tp 0 --locprob 0.75 --uniqueness 0 --multidir . \
  --filelist $work_dir/samples/filelist_ionquant.txt \
  --modlist  $work_dir/samples/modmasses_ionquant.txt

runSEMQuant.sh in SEMQuant-DDA/ shows the same sequence as a single driver script.

SEMQuant-Astral

Astral workflow: raw conversion → Sipros database search → Percolator rescoring → ProteinProphet inference → Philosopher FDR control → spectral library generation → DIA-NN quantification.

1. Set up the working directory

export work_dir=/PATH_TO_YOUR_WORK_DIR
export astral_dir=$tool_dir/SEMQuant-Astral

mkdir -p $work_dir/{raw,samples,results}

2. Download example raw files

cd $work_dir/raw
# three-species Astral benchmark, E45 condition (PXD046444)
wget ftp://ftp.ebi.ac.uk/pride/archive/projects/PXD046444/20230324_OLEP08_200ng_30min_E45H50Y5*.raw

This yields three sample raw files.

3. Convert raw to indexed mzML

conda activate mono
# -f=2 indexed mzML output, -L=2 keep MS2 only
mono ThermoRawFileParser.exe -d=$work_dir/raw -o=$work_dir/raw -f=2 -L=2

4. Database search

Uses the same search engine as the DDA pipeline ($tool_dir/SEMQuant-DDA/Sipros_OpenMP), on a single MS2 file (-f) or a whole directory (-w):

export OMP_NUM_THREADS=24
export sipros=$tool_dir/SEMQuant-DDA/Sipros_OpenMP

# single MS2 file
$sipros -o $work_dir/results -f $work_dir/raw/YOUR_MS2.mzML \
        -c $tool_dir/SEMQuant-DDA/configs/SiprosConfig.cfg

# all MS2 files in a directory
$sipros -o $work_dir/results -w $work_dir/raw \
        -c $tool_dir/SEMQuant-DDA/configs/SiprosConfig.cfg

5. Generate PSM features for Percolator

An R script converts the Sipros PSM output into a Percolator .pin feature file. Run it once per sample to produce <sample>.pin.

6. Rescore PSMs with Percolator

export sample=20230324_OLEP08_200ng_30min_E45H50Y5_180K_2Th3p5ms_02

$astral_dir/percolator_3_6_4/linux/percolator \
    --only-psms --no-terminate --post-processing-tdc --num-threads 23 \
    --results-psms       ${sample}_percolator_target_psms.tsv \
    --decoy-results-psms ${sample}_percolator_decoy_psms.tsv \
    --protein-decoy-pattern rev_ \
    ${sample}.pin

7. Convert Percolator output for protein inference

python $astral_dir/SE_IQ_Astral/Percolator2PeptideProphet.py \
    $work_dir/results/E45/01/${sample}.target.Spe2Pep.txt \
    $work_dir/results/E45/01/${sample}.pin \
    $work_dir/results/E45/01/ \
    ${sample}

This writes interact-<sample>.pep.xml files for ProteinProphet.

8. Protein inference

List the per-run interact-*.pep.xml files in a file list (see SEMQuant-Astral/filelist_proteinprophet.txt for the expected format), then:

$astral_dir/philosopher_v5.1.0_linux_amd64/philosopher proteinprophet \
    --maxppmdiff 2000000 --output combined \
    $work_dir/E45/filelist_proteinprophet.txt

9. FDR control

export philosopher=$astral_dir/philosopher_v5.1.0_linux_amd64/philosopher
export db=$work_dir/2024-09-22-decoys-mix_HYE.fasta.fas

$philosopher database --annotate $db --prefix rev_

$philosopher filter --picked --razor --prot 0.01 --minPepLen 8 --tag rev_ \
    --pepxml  $work_dir/E45 \
    --protxml $work_dir/E45/combined.prot.xml

10. Generate the spectral library

Uses FragPipe's EasyPQP-based library builder (obtain speclib/gen_con_spec_lib.py and unimod.xml from a FragPipe installation):

python -u $speclib_dir/gen_con_spec_lib.py \
    $db $work_dir/E45 unused $work_dir/E45 True unused use_easypqp noiRT;noIM 16 \
    "--unimod $speclib_dir/unimod.xml --max_delta_unimod 0.02 --max_delta_ppm 15.0 --fragment_types [\'b\',\'y\',]" \
    "--rt_lowess_fraction 0.0" \
    delete_intermediate_files \
    $work_dir/E45/filelist_speclibgen.txt

11. Quantify with DIA-NN

Unzip the bundled DIA-NN 1.8.2 beta 8 build first:

git lfs pull                                  # if not already fetched
unzip $astral_dir/linux_diann.zip -d $astral_dir/diann

$astral_dir/diann/diann-1.8.1.8 \
    --lib library.tsv --threads 15 --verbose 1 \
    --out diann-output/report.tsv \
    --qvalue 0.01 --matrix-qvalue 0.01 --matrices \
    --no-prot-inf --smart-profiling --no-quant-files \
    --peak-center --no-ifs-removal --report-lib-info \
    --cfg $work_dir/E45/filelist_diann.txt

Citation

If you use SEMQuant in your research, please cite:

Zhang B., Feng S., Xiong Y., Pan C., Guo X. SEMQuant: A computational platform for accurate and comprehensive quantitative metaproteomics analysis. Journal of Computer Science and Technology, 2026, 41(3): 1087–1100. https://doi.org/10.1007/s11390-026-5122-3

@article{zhang2026semquant,
  title   = {SEMQuant: A Computational Platform for Accurate and Comprehensive
             Quantitative Metaproteomics Analysis},
  author  = {Zhang, Bailu and Feng, Shichao and Xiong, Yi and Pan, Chongle and Guo, Xuan},
  journal = {Journal of Computer Science and Technology},
  volume  = {41},
  number  = {3},
  pages   = {1087--1100},
  year    = {2026},
  doi     = {10.1007/s11390-026-5122-3},
  url     = {https://doi.org/10.1007/s11390-026-5122-3}
}

Name		Name	Last commit message	Last commit date
SEMQuant-Astral		SEMQuant-Astral
SEMQuant-DDA		SEMQuant-DDA
.gitattributes		.gitattributes
README.md		README.md

Navigation Menu

Repository files navigation

SEMQuant

Table of Contents

Repository layout

Requirements

Conventions

SEMQuant-DDA

1. Set up the working directory

2. Download example raw files

3. Convert raw to FT2

4. Generate the target–decoy database

5. Database search

6. Split results into per-sample folders

7. Filter PSMs and assemble proteins

8. Convert results for IonQuant

9. Quantify with IonQuant

SEMQuant-Astral

1. Set up the working directory

2. Download example raw files

3. Convert raw to indexed mzML

4. Database search

5. Generate PSM features for Percolator

6. Rescore PSMs with Percolator

7. Convert Percolator output for protein inference

8. Protein inference

9. FDR control

10. Generate the spectral library

11. Quantify with DIA-NN

Citation

About

Releases

Packages

Contributors

Languages

Latest commit

History

Folders and files

Repository files navigation

SEMQuant

Table of Contents

Repository layout

Requirements

Conventions

SEMQuant-DDA

1. Set up the working directory

2. Download example raw files

3. Convert raw to FT2

4. Generate the target–decoy database

5. Database search

6. Split results into per-sample folders

7. Filter PSMs and assemble proteins

8. Convert results for IonQuant

9. Quantify with IonQuant

SEMQuant-Astral

1. Set up the working directory

2. Download example raw files

3. Convert raw to indexed mzML

4. Database search

5. Generate PSM features for Percolator

6. Rescore PSMs with Percolator

7. Convert Percolator output for protein inference

8. Protein inference

9. FDR control

10. Generate the spectral library

11. Quantify with DIA-NN

Citation

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages