Merge pull request #137 from compomics/feature/non-tryptic

Add non-tryptic peptide and immunopeptide prediction models

.gitattributes

@@ -0,0 +1 @@
+ms2pip/models/**/*.c filter=lfs diff=lfs merge=lfs -text

.github/workflows/build_and_publish.yml

@@ -10,6 +10,8 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v2
+        with:
+          lfs: 'true'
       - name: Set up Python
         uses: actions/setup-python@v2
         with:
@@ -36,6 +38,8 @@ jobs:
         os: [ubuntu-18.04, macos-latest]
     steps:
       - uses: actions/checkout@v2
+        with:
+          lfs: 'true'
       - uses: actions/setup-python@v2
         name: Install Python
         with:

.github/workflows/test.yml

@@ -13,6 +13,8 @@ jobs:
 
     steps:
     - uses: actions/checkout@v2
+      with:
+        lfs: 'true'
     - name: Set up Python ${{ matrix.python-version }}
       uses: actions/setup-python@v2
       with:

.gitignore

@@ -4,6 +4,7 @@
 data/
 *_pyx.c
 *_pyx_*.c
+ms2pip/models_xgboost/*.xgboost
 
 # Pytest
 .pytest_cache/

README.md

@@ -286,26 +286,30 @@ next to the predictions for singly charged b- and y-ions.
 
 | Model | Fragmentation method | MS² mass analyzer | Peptide properties |
 | - | - | - | - |
-| HCD | HCD | Orbitrap | Tryptic digest |
+| HCD2019 | HCD | Orbitrap | Tryptic digest |
+| HCD2021 | HCD | Orbitrap | Tryptic/ Chymotrypsin digest |
 | CID | CID | Linear ion trap | Tryptic digest |
 | iTRAQ | HCD | Orbitrap | Tryptic digest, iTRAQ-labeled |
 | iTRAQphospho | HCD | Orbitrap | Tryptic digest, iTRAQ-labeled, enriched for phosphorylation |
 | TMT | HCD | Orbitrap | Tryptic digest, TMT-labeled |
 | TTOF5600 | CID | Quadrupole Time-of-Flight | Tryptic digest |
 | HCDch2 | HCD | Orbitrap | Tryptic digest |
 | CIDch2 | CID | Linear ion trap | Tryptic digest |
+| Immuno-HCD | HCD | Orbitrap | Immunopeptides |
 
 ### Models, version numbers, and the train and test datasets used to create each model
 
 | Model | Current version | Train-test dataset (unique peptides) | Evaluation dataset (unique peptides) | Median Pearson correlation on evaluation dataset |
 | - | - | - | - | - |
-| HCD | v20190107 | [MassIVE-KB](https://doi.org/10.1016/j.cels.2018.08.004) (1 623 712) | [PXD008034](https://doi.org/10.1016/j.jprot.2017.12.006) (35 269) | 0.903786 |
+| HCD2019 | v20190107 | [MassIVE-KB](https://doi.org/10.1016/j.cels.2018.08.004) (1 623 712) | [PXD008034](https://doi.org/10.1016/j.jprot.2017.12.006) (35 269) | 0.903786 |
 | CID | v20190107 | [NIST CID Human](https://chemdata.nist.gov/) (340 356) | [NIST CID Yeast](https://chemdata.nist.gov/) (92 609) | 0.904947 |
 | iTRAQ | v20190107 | [NIST iTRAQ](https://chemdata.nist.gov/) (704 041) | [PXD001189](https://doi.org/10.1182/blood-2016-05-714048) (41 502) | 0.905870 |
 | iTRAQphospho | v20190107 | [NIST iTRAQ phospho](https://chemdata.nist.gov/) (183 383) | [PXD001189](https://doi.org/10.1182/blood-2016-05-714048) (9 088) | 0.843898 |
 | TMT | v20190107 | [Peng Lab TMT Spectral Library](https://doi.org/10.1021/acs.jproteome.8b00594) (1 185 547) | [PXD009495](https://doi.org/10.15252/msb.20188242) (36 137) | 0.950460 |
 | TTOF5600 | v20190107 | [PXD000954](https://doi.org/10.1038/sdata.2014.31) (215 713) | [PXD001587](https://doi.org/10.1038/nmeth.3255) (15 111) | 0.746823 |
 | HCDch2 | v20190107 | [MassIVE-KB](https://doi.org/10.1016/j.cels.2018.08.004) (1 623 712) | [PXD008034](https://doi.org/10.1016/j.jprot.2017.12.006) (35 269) | 0.903786 (+) and 0.644162 (++) |
 | CIDch2 | v20190107 | [NIST CID Human](https://chemdata.nist.gov/) (340 356) | [NIST CID Yeast](https://chemdata.nist.gov/) (92 609) | 0.904947 (+) and 0.813342 (++) |
+| HCD2021 | v20210416 | [Combined dataset] (520 579) | [PXD008034](https://doi.org/10.1016/j.jprot.2017.12.006) (35 269)  | 0.932361
+| Immuno-HCD | v20210316 | [Combined dataset] (460191) | [PXD005231 (HLA-I)](https://doi.org/10.1101/098780) (46 753) <br>[PXD020011 (HLA-II)](https://doi.org/10.3389/fimmu.2020.01981 ) (23 941) | 0.963736<br>0.942383
 
 To train custom MS²PIP models, please refer to [Training new MS²PIP models](http://compomics.github.io/projects/ms2pip_c/wiki/Training-new-MS2PIP-models.html) on our Wiki pages.

ms2pip/__main__.py

@@ -6,21 +6,23 @@
 from ms2pip.config_parser import ConfigParser
 from ms2pip.exceptions import (FragmentationModelRequiredError,
                                InvalidModificationFormattingError,
-                               InvalidPEPRECError,
+                               InvalidPEPRECError, InvalidXGBoostModelError,
                                NoValidPeptideSequencesError,
                                UnknownFragmentationMethodError,
                                UnknownModificationError,
-                               UnknownOutputFormatError)
+                               UnknownOutputFormatError,
+                               InvalidXGBoostModelError,
+                               EmptySpectrumError)
 from ms2pip.ms2pipC import MODELS, MS2PIP, SUPPORTED_OUT_FORMATS
 
 
 def print_logo():
-    logo = """
- __  __ ___  __ ___ ___ ___  
-|  \/  / __||_ ) _ \_ _| _ \ 
-| |\/| \__ \/__|  _/| ||  _/ 
-|_|  |_|___/   |_| |___|_|   
-                             
+    logo = r"""
+ __  __ ___  __ ___ ___ ___
+|  \/  / __||_ ) _ \_ _| _ \
+| |\/| \__ \/__|  _/| ||  _/
+|_|  |_|___/   |_| |___|_|
+
 by CompOmics
 sven.degroeve@ugent.be
 ralf.gabriels@ugent.be
@@ -155,6 +157,12 @@ def main():
     except FragmentationModelRequiredError:
         root_logger.error("Please specify model in config file.")
         sys.exit(1)
+    except InvalidXGBoostModelError:
+        root_logger.error(f"Could not download XGBoost model properly\nTry manual download")
+        sys.exit(1)
+    except EmptySpectrumError:
+        root_logger.error("Provided MGF file cannot contain empty spectra")
+        sys.exit(1)
 
 
 if __name__ == "__main__":

ms2pip/cython_modules/ms2pip_peaks_c.c

@@ -9,23 +9,23 @@
 #include "ms2pip_features_c_catboost.c"
 
 #include "../models/CID.h"
-#include "../models/HCD.h"
+#include "../models/HCD-2019.h"
 #include "../models/TTOF5600.h"
 #include "../models/TMT.h"
 #include "../models/iTRAQ.h"
 #include "../models/iTRAQphospho.h"
 
 float membuffer[10000];
-float ions[2000]; 
-float mzs[2000]; 
+float ions[2000];
+float mzs[2000];
 float predictions[1000];
 
 struct annotations{
 	float* peaks;
 	float* msms;
 };
 typedef struct annotations annotations;
- 
+
 //compute feature vector from peptide + predict intensities
 float* get_p_ms2pip(int peplen, unsigned short* peptide, unsigned short* modpeptide, int charge, int model_id, int ce)
 	{
@@ -80,7 +80,7 @@ float* get_p_ms2pip(int peplen, unsigned short* peptide, unsigned short* modpept
 			predictions[2*(peplen-1)-i-1] = score_iTRAQphospho_Y(v+1+(i*fnum))+0.5;
 		}
 	}
-	
+
 	// EThcD
 	// else if (model_id == 6) {
 	// 	for (i=0; i < peplen-1; i++) {
@@ -270,7 +270,7 @@ annotations get_t_ms2pip_all(int peplen, unsigned short* modpeptide, int numpeak
 	//  fprintf(stderr,"m %f\n",msms[i]);
 	//}
 
-	for (i=0; i < 18*(peplen-1); i++) { // 2*9 iontypes: b: a -H2O -NH3 b c y: -H2O z y x 
+	for (i=0; i < 18*(peplen-1); i++) { // 2*9 iontypes: b: a -H2O -NH3 b c y: -H2O z y x
 		ions[i] = -9.96578428466; //HARD CODED!!
 		mzs[i] = 0; //HARD CODED!!
 	}
@@ -489,14 +489,14 @@ annotations get_t_ms2pip_all(int peplen, unsigned short* modpeptide, int numpeak
 			mem_pos += 1;
 		}
 	}
-	
-	//for (i=0; i < 18*(peplen-1); i++) { // 2*9 iontypes: b: a -H2O -NH3 b c y: -H2O z y x 
+
+	//for (i=0; i < 18*(peplen-1); i++) { // 2*9 iontypes: b: a -H2O -NH3 b c y: -H2O z y x
 	//    fprintf(stderr,"%f ",ions[i]); //HARD CODED!!
 	//}
 	//fprintf(stderr,"\n");
 
 	struct annotations r = {ions,mzs};
-	
+
 	return r;
 }
 

ms2pip/cython_modules/ms2pip_pyx.pyx

@@ -14,7 +14,7 @@ cdef extern from "ms2pip_peaks_c.c":
 		float* msms
 
 	void init_ms2pip(char* amino_masses_fname, char* modifications_fname, char* modifications_fname_sptm)
-	
+
 	unsigned int* get_v_ms2pip(int peplen, unsigned short* peptide, unsigned short* modpeptide, int charge)
 	unsigned int* get_v_ms2pip_ce(int peplen, unsigned short* peptide, unsigned short* modpeptide, int charge, int ce)
 	unsigned int* get_v_ms2pip_old(int peplen, unsigned short* peptide, unsigned short* modpeptide, int charge)
@@ -134,10 +134,10 @@ def get_targets_all(np.ndarray[unsigned short, ndim=1, mode="c"] modpeptide,
 	results = get_t_ms2pip_all(len(modpeptide)-2, &modpeptide[0], len(peaks), &msms[0], &peaks[0], fragerror)
 	result_peaks = []
 	for i in range(NUM_ION_TYPES_MAPPING[peaks_version]*(len(modpeptide)-3)):
-		result_peaks.append(results.peaks[i])   
+		result_peaks.append(results.peaks[i])
 	result_mzs = []
 	for i in range(NUM_ION_TYPES_MAPPING[peaks_version]*(len(modpeptide)-3)):
-		result_mzs.append(results.msms[i])  
+		result_mzs.append(results.msms[i])
 	#print(result_parsed)
 	#print()
 	return (result_mzs,result_peaks)

ms2pip/exceptions.py

@@ -40,3 +40,11 @@ class InvalidModificationFormattingError(Exception):
 
 class InvalidAminoAcidError(Exception):
     pass
+
+
+class EmptySpectrumError(Exception):
+    pass
+
+
+class InvalidXGBoostModelError(Exception):
+    pass