metaExpertPro is a computational software for the analysis of DDA-MS and DIA-MS hybrid-based metaproteomic data. It provides a complete pipeline for peptide and protein measurement, functional and taxonomic annotation, and generation of quantitative matrices for both microbiota and hosts. It is compatible with DDA-MS and DIA-MS data from Thermo Fisher Orbitrap ( .mzML format) or Bruker (.d format) mass spectrometers.
The metaExpertPro software is supported by Docker, which enables running under both Windows and Linux operating systems.
The Docker container includes Java (11.0.13), Python (v3.9.18), easypqp (v0.1.35), DIA-NN (v1.8), Unipept (v3.1.0), eggnog-mapper (v2.1.5), and all the necessary environments of these software tools.
The Docker container doesn't include FragPipe software, but it provides a 'split-database' workflow and running commands for FragPipe v20.0.
To use FragPipe v20.0 for DDA-MS-based spectral library generation, download it from this link (https://github.com/Nesvilab/FragPipe/releases) and specify the FragPipe path during the process.
It is recommended to have at least 128 GB RAM and 50 GB storage space for one DDA-MS raw data (.d or .mzML) based metaExpertPro analysis.
First, download the metaExpertPro container from Docker Hub.
$ docker pull guomics2017/metaexpertpro:v2.3
Part 1: run metaExpertPro for DDA-MS-based spectral library generation and DIA-MS-based peptide and protein quantification.
The following folders are required:
- A folder as the total directory for metaExpertPro (metaEx).
- A folder for DDA raw data input files (metaEx/DDAraw).
- A folder for DIA raw data input files (metaEx/DIAraw).
- A folder for protein sequence database input files (metaEx/fasta)
- A folder for analysis results (metaEx/Results)
The file format is required as follows:
- DDA and DIA raw data input file formats are .d or .mzML.
- Database input file is required in .fasta format.
docker run -it --rm -u $(id -u):$(id -g) guomics2017/metaexpertpro:v2.3 sh /metaEx/src/00.DDAspectrallib/00.DDA.DIA.sh --help
The users can choose whether or not to run FragPipe and DIA-NN via the command line.
For DDA-MS database search:
- True precursor mass tolerance (unit ppm): 20
- Fragment mass tolerance units (0 for Da, 1 for ppm): 1
- Fragment mass tolerance: 20
- Minimum length of peptides to be generated during in-silico digestion: 7
- Maximum length of peptides to be generated during in-silico digestion: 50
- Allowed number of missed cleavages per peptide: 2
- Cleavage site: Trypsin
- Threads: 20
For DIA-MS database search:
- True precursor mass tolerance (unit ppm): 10
- Fragment mass tolerance (unit ppm): 10
- Allowed number of missed cleavages per peptide: 1
- Threads: 20
docker run -it --rm \
-u $(id -u):$(id -g) \
-v /workdir/metaEx/DDAraw/:/metaEx/DDAraw/ \
-v /workdir/metaEx/DIAraw/:/metaEx/DIAraw/ \
-v /workdir/metaEx/fasta/:/metaEx/fasta/ \
-v /workdir/metaEx/Results/:/metaEx/Results/ \
-v /path/to/fragpipe/:/metaEx/software/fragpipe20.0/ \
guomics2017/metaexpertpro:v2.3 sh /metaEx/src/00.DDAspectrallib/00.DDA.DIA.sh \
--total_dir /metaEx --project_name xxx --fragpipe_switch xxx --diann_switch xxx --fasta_name xxx --dia_threads xxx \
--fragpipe_path /metaEx/software/fragpipe20.0 --db_split xxx
- DDA-MS-based spectral library: metaEx/Results/00.DDAspectrallib/library.tsv
- DDA-MS-based protein sequences: metaEx/Results/00.DDAspectrallib/protein.fas
- DIA-MS-based peptide and protein quantitative matrices: metaEx/Results/01.DIAquant
Part 2: run metaExpertPro for functional and taxonomic annotation and quantitative matrices generation.
- For eggnog-mapper-based eggnog annotation:
For conducting eggnog annotation via eggnog-mapper, the essential eggnog-mapper database files, namely eggnog.db and eggnog_proteins.dmnd, are necessary. You can obtain the compressed files by accessing the following link (https://pan.baidu.com/s/11ZTPnbgz5p2c75W-ATJ8JA) on Baidu Cloud Drive (with the extraction code being yosa). Subsequently, these files need to be decompressed and positioned within the directory path metaEx/software/eggnog-mapper/eggnog-mapper-data/.
Note: The sizes of the eggnog.db and eggnog_proteins.dmnd files are approximately 42 GB.
- For GhostKOALA-based KO annotation:
The GhostKOALA-based KO annotation can only be done through the webserver (https://www.kegg.jp/ghostkoala/) The users need to upload the .fas file in the folder metaEx/Results/00.DDAspectrallib/protein.fas generated from the Part 1 run to the GhostKOALA.
Then, place the results of GhostKOALA in the folder metaEx/Results
- A folder for sample name, batch ID, and label (metaEx/sampleLabel):
Sample label input file is required as .csv format and the example content is shown in the example folder.
- It is recommended to have at least 32 GB RAM for annotation analysis.
docker run -it --rm -u $(id -u):$(id -g) guomics2017/metaexpertpro:v2.3 sh /metaEx/src/02.Annotation/01.annotation.sh --help
The users can choose whether or not to run Unipept, eggnog-mapper, and KEGG annotation via the command line.
- threads: 20
- sample label file: /metaEx/sampleLabel/my_project_sample_label.csv
docker run -it --rm \
-u $(id -u):$(id -g) \
-v /workdir/metaEx/sampleLabel/:/metaEx/sampleLabel/ \
-v /workdir/metaEx/Results/:/metaEx/Results/ \
-v /workdir/metaEx/software/eggnog-mapper/eggnog-mapper-data/:/metaEx/software/eggnog-mapper/eggnog-mapper-data/ \
guomics2017/metaexpertpro:v2.3 sh /metaEx/src/02.Annotation/01.annotation.sh --total_dir /metaEx --project_name xxx --sample_label /metaEx/sampleLabel/xxx \
--database xxx --input_pr_mat_name xxx --input_pg_mat_name xxx --anno_threads xxx --unipept_switch xxx --eggnog_switch xxx --kegg_switch xxx
All the matrices are located in the metaEx/Results/02.Annotation/07.matrix. The folder includes the following folders:
- all/: all the samples
- sample/: subjective samples
- qc/: inter-batch biological replicates
- pool/: inter-batch technical replicates
- biorep/: intra-batch biological replicates
- techrep/: intra-batch technical replicates
Each of the above folders contains peptide, protein, COG, COG category, KO, KO category, and taxa quantitative matrices for the corresponding samples.