Annotate phosphoproteomic dataset

In many cases, you may want to append information from ProteomeScout to a particular dataset. This is easily done through the Proteomics class, which is designed to read in UniProt accessions and a peptide and find information associated with each peptide.

In order for this to work, the annotation class expects the peptide to be formatted such that any phosphorylation sites are annotated as lowercase, and all other residues are formatted as uppercase. For example, the following peptide has

Load dataset

First, load the dataset as a dataframe. Notice the format of the peptides in the ‘pep’ column, with a lowercased phosphorylation site. If there are multiple phosphorylated sites in the peptide, there should can be multiple lowercased residues.

import pandas as pd

df = pd.read_csv('example_dataset_for_annotation.csv')
df.head()

	acc	Protein name	Protein symbol	Phosphorylation site	modification	pep	data:time(hours):0	data:time(hours):0.5	data:time(hours):1	data:time(hours):2	data:time(hours):4	data:time(hours):8	data:time(hours):24
0	Q9C0C2	182 kDa tankyrase-1-binding protein.	TNKS1BP1	S672	phosphorylation	TEAQDLCRAsPEPPGPESSSR	1.05	0.75	0.59	0.89	0.95	0.89	1.0
1	Q9C0C2	182 kDa tankyrase-1-binding protein.	TNKS1BP1	S691	phosphorylation	WLDDLLAsPPPSGGGAR	0.56	0.66	0.63	0.58	0.48	0.64	1.0
2	P17980	26S protease regulatory subunit 6A	PSMC3	Y132	phosphorylation	QTyFLPVIGLVDAEK	0.11	0.09	0.11	0.12	0.14	0.04	1.0
3	O15530	3-phosphoinositide-dependent protein kinase 1	PDPK1	S241	phosphorylation	ANsFVGTAQYVSPELLTEK	0.64	0.81	0.58	0.98	0.62	0.66	1.0
4	P23396	40S ribosomal protein S3.	RPS3	T220/ T221*	phosphorylation	DEILPtTPISEQK	0.73	0.64	0.59	0.56	0.61	0.58	1.0

Once data has been loaded, you simply need to pass the dataset to the ProteomicDataset class, indicating which columns contain the SwissProt accession and which columns contain your formatted peptide.

from proteomeScoutAPI import ProteomicDataset

#initialize ProteomicsDataset object
dataset = ProteomicDataset(df, accession_col='acc', peptide_col='pep', find_site=True, GO_terms=True)

#annotate the dataset
dataset.annotate_dataset()

Annotation information has been appended to the dataset DataFrame as new columns.

# to save the dataset to a csv file, uncomment the following line
# dataset.dataset.to_csv('annotated_dataset.csv', index=False)

dataset.dataset
# print the new columns that were added to the dataset
orig_colums = df.columns
new_columns = dataset.dataset.columns
print("These columns were added to the dataset:")
for col in new_columns:
    if col not in orig_colums:
        print(col)

These columns were added to the dataset:
gene_name
domains
domain_architecture
GO_terms
aligned_peps
modification_sites
documented_modification
documented_modification_details
site_in_domain:name
site_in_domain:interpro
site_in_macro
site_in_structure
site_in_activation_loop
site_in_exon
SpY-C Prediction
SpY-C Prediction Probability
PScout_Errors

This will append two types of information, either based on the entire gene (gene name, domains, GO terms) or the individual ptms (position, modification, if in domain or macromolecular structure)

Gene level information will look like this:

dataset.dataset[['gene_name', 'domains', 'domain_architecture', 'GO_terms', 'PScout_Errors']].head()

	gene_name	domains	domain_architecture	GO_terms	PScout_Errors
0	TNKS1BP1	Tankyrase-bd_C:1554:1726:IPR032764	Tankyrase-bd_C	GO:0071479-P:cellular response to ionizing rad...	NaN
1	TNKS1BP1	Tankyrase-bd_C:1554:1726:IPR032764	Tankyrase-bd_C	GO:0071479-P:cellular response to ionizing rad...	NaN
2	PSMC3	Prot_ATP_ID_OB_2nd:112:165:IPR032501;AAA+_ATPa...	Prot_ATP_ID_OB_2nd~AAA+_ATPase~AAA_lid_3	GO:0001824-P:blastocyst development;  GO:00439...	NaN
3	PDPK1	Prot_kinase_dom:82:342:IPR000719;PDK1-typ_PH:4...	Prot_kinase_dom~PDK1-typ_PH	GO:0030036-P:actin cytoskeleton organization; ...	NaN
4	RPS3	KH_dom_type_2:20:93:IPR004044;Ribosomal_uS3_C:...	KH_dom_type_2~Ribosomal_uS3_C	GO:0006915-P:apoptotic process;  GO:0006284-P:...	NaN

Each column contains:

• gene_name: gene name associated with SwissProt accession

• domains: domain information associated with the protein, with each entry separated by a semicolon. Domains are structured like “domain_name:start:stop:interpro_ID”

• domain_architecture: Presents each domain in the order they appear in the protein

• GO_terms: gene ontology terms associated with the gene

PTM-level information will look like this:

dataset.dataset[['aligned_peps', 'modification_sites', 'documented_modification', 'site_in_domain:name', 'site_in_domain:interpro', 'site_in_macro', 'site_in_structure', 'site_in_activation_loop', 'SpY-C Prediction']].head(n=10)

	aligned_peps	modification_sites	documented_modification	site_in_domain:name	site_in_domain:interpro	site_in_macro	site_in_structure	site_in_activation_loop	SpY-C Prediction
0	AQDLCRAsPEPPGPE	S672	1			Acidic;Disordered
1	WLDDLLAsPPPSGGG	S691	1			Acidic;Disordered
2	KTSTRQTyFLPVIGL	Y132	1	Prot_ATP_ID_OB_2nd	IPR032501		STRAND		binder
3	SKQARANsFVGTAQY	S241	1	Prot_kinase_dom	IPR000719			HIGH
4	PKDEILPtTPISEQK	T220	1			Disordered
5	RTVIIEQsWGSPKVT	S67	1
6	KVEAKEEsEESDEDM;AKEESEEsDEDMGFG	S304;S307	1;1			Disordered;Acidic residues;Disordered;Acidic r...			;
7	VEAKKEEsEESDDDM;KKEESEEsDDDMGFG	S101;S104	1;1			Disordered;Disordered			;
8	VEAKKEEsEESDDDM	S101	1			Disordered
9	LAALGGNsSPSAKDI	S16	1

Each column indicates:

• aligned_peps: sequence surrounding phosphrylation sites (+/- 7 amino acids)

• modification_sites: indicates the amino acid and position of modification sites found. Comma separated if more than 1 modification was in the peptide

• documented_modification: indicates whether the specific site is found in ProteomeScout (1) or not (0) as a documented modification

• documented_modification_details: indicates the type of modifications seen at that site and the experiment IDs in (list) format.

• site_in_domain:name: protein domains name from Interpro that contain the modification site

• site_in_domain:interpro: Interpro ID for any domain that a modification site exists in

• site_in_macro: macromolecular structures that contain the phosphosite

• site_in_structure: secondary structures associated with phosphosite

• site_in_activation_loop: indicates if this is an activation loop of a kinase and indicates the confidence in the activation loop (HIGH - well annotated loop, MEDIUM - at least part of the anchoring sequence found and good alignment)

• site_in_exon: gives the exon ID that a modification falls in

• SpY-C Prediction: If a tyrosine phosphorylation site this gives the binding class (binder, nonbinder) for recognition by an SH2 domain

• SpY-C Prediction Probability: If a tyrosine phosphroylation site that has a SpY-C prediction, this is the probability score.