Inspecting number of PTMs with annotation information available

As described in Running PTM-POSE section, PTM-POSE provides various options for annotating functional information for PTMs, coming from various databases. However, PTM functional information is inherently sparse, and so most annotations will only provide information on a handful of PTMs. For this reason, it can be useful to probe how many PTMsTo better understand the types of annotations that are available, as well as the number of PTMs that have an annotation of that type. This can be done using the analyze function in PTM-POSE.

Note: This examples assumes that you have already run the PTM-POSE pipeline and have at annotated PTMs with at least one layer of information.

from ptm_pose import helpers
from ptm_pose.analyze import annotations
import matplotlib.pyplot as plt

#load example differential inclusion data
spliced_ptms = helpers.load_example_data(spliced_ptms = True)

available_annotations = annotations.get_available_annotations(spliced_ptms)
available_annotations

	Database	Annotation Type	Appended to PTM data?
0	iKiP	Enzyme	No
1	PhosphoSitePlus	Enzyme	No
2	PhosphoSitePlus	Disease	No
3	PhosphoSitePlus	Interactions	No
4	PhosphoSitePlus	Function	No
5	PhosphoSitePlus	Process	No
6	PhosphoSitePlus	Perturbation	No
7	PTMsigDB	Perturbation-DIA	No
8	PTMsigDB	Perturbation-DIA2	No
9	PTMsigDB	Pathway-NetPath	No
10	PTMsigDB	Pathway-WikiPathways	No
11	PTMsigDB	Perturbation-PRM	No
12	PTMsigDB	Pathway-BI	No
13	DEPOD	Enzyme	No
14	RegPhos	Enzyme	No

You can also visualize the number of PTMs with annotation information available for each annotation type using a bar plot.

annotations.plot_available_annotations(spliced_ptms)
plt.tight_layout()
plt.show()

As you can, see there are only a few PTMs from each annotation that have available information, with the most being 9 PTMs out of the 184 differentially included sites having been associated with a biological process. While this this should be taken into consideration when analyzing these annotations, we can glean some useful information and identify potentially interesting proteins/sites to dig deeper into. Let’s look at the PTMs that have been associated with a biological process:

ptms_with_annotation, annotation_counts = annotations.get_ptm_annotations(spliced_ptms, database = "PhosphoSitePlus", annotation_type = 'Process')
print('Specific PTMs with annotation:')
ptms_with_annotation

3 PTMs removed due to insignificant splice event (p < 0.05, dpsi >= 0.1): (33.33%)
Final number of PTMs to be assessed: 6
Specific PTMs with annotation:

	Gene	UniProtKB Accession	Residue	PTM Position in Isoform	Modification Class	PhosphoSitePlus:Function	dPSI	Significance	Impact
0	CEACAM1	P13688	S	461.0	Phosphorylation	activity, inhibited	0.525	1.73943268451e-09	Included
1	SPHK2	Q9NRA0	S	419.0	Phosphorylation	intracellular localization	0.253	0.0129400018182	Included
2	SPHK2	Q9NRA0	S	421.0	Phosphorylation	intracellular localization	0.253	0.0129400018182	Included
3	TSC2	P49815	S	981.0	Phosphorylation	molecular association, regulation;intracellula...	-0.219	4.18472157275e-05	Excluded
4	YAP1	P46937	K	342.0	Ubiquitination	protein degradation	-0.188;-0.161	4.17884655686e-07;0.000211254197372	Excluded

We can also look at the number of PTMs associated with each annotation:

print('Number of PTMs associated with each annotation:')
annotation_counts

Number of PTMs associated with each annotation:

	All Impacted	Included	Excluded
PhosphoSitePlus:Function
intracellular localization	3	2.0	1
activity, inhibited	2	1.0	1
molecular association, regulation	1	0.0	1
protein degradation	1	0.0	1

Note: you could also do this analysis for altered flanking sequences by replacing spliced_ptms with altered_flanks in the above code.