1. Install HELP from GitHub (and Karateclub)

Skip this cell if you already have installed HELP and Karateclub.

!pip install git+https://github.com/giordamaug/HELP.git
!pip install -q karateclub

2. Download the PPI input file

The file describing the PPI network for the Kidney tissue can be found on GitHub.

!wget https://raw.githubusercontent.com/giordamaug/HELP/main/data/Kidney_PPI.csv

3. Load the PPI input file

The PPI network is given as a csv file based on the data downloaded from the Integrated Interaction Database (IID), where each row represents one of the edges, and the three columns represent: + Column A: gene name of the edge source; + Column B: gene name of the edge target; + Column combined_score: attribute providing the edge weight, which can be 1 if supported by one piece of evidence, 2 if supported by two or 3 if supported by three.

import pandas as pd
from HELPpy.preprocess.embedding import PPI_embed
df_net = pd.read_csv('Kidney_PPI.csv')

4. Compute the embedding

Compute the result using a graph embedding method (here method="Node2Vec") to produce embedding vectors of default length (128).

df_embed = PPI_embed(df_net, method="Node2Vec", verbose=True)

Please be aware that this will take almost 2 hours in sequential execution.

5. Save and show the embedding

The embedding result is a DataFrame with rows the gene names and columns the components of the 128-sized vector representing each node/gene in the PPI.

df_embed.to_csv('Kidney_EmbN2V_128.csv')
df_embed

	Node2Vec_0	Node2Vec_1	Node2Vec_2	Node2Vec_3	Node2Vec_4	Node2Vec_5	Node2Vec_6	Node2Vec_7	Node2Vec_8	Node2Vec_9	...	Node2Vec_118	Node2Vec_119	Node2Vec_120	Node2Vec_121	Node2Vec_122	Node2Vec_123	Node2Vec_124	Node2Vec_125	Node2Vec_126	Node2Vec_127
(clone tec14)	0.065830	0.015107	-0.169474	-0.017198	0.085214	0.066795	0.021665	-0.033675	0.127271	0.086943	...	-0.139271	-0.192168	-0.024476	-0.059446	0.030652	0.011218	-0.095148	-0.065400	-0.069221	0.019122
100 kDa coactivator	-0.120776	0.333710	-0.264869	0.258195	-0.137268	0.223037	0.344388	-0.058348	0.050687	0.190354	...	-0.447794	-0.207273	-0.021812	0.102079	0.379066	0.227051	-0.299870	0.093029	0.420842	-0.299310
14-3-3 tau splice variant	-0.207174	0.489269	0.060112	0.033272	-0.582095	0.089639	0.183833	-0.331528	-0.033732	-0.316844	...	-0.135543	-0.491769	0.059879	0.572159	-0.167333	-0.774573	-0.329807	0.241468	-0.139246	0.181745
3'-phosphoadenosine-5'-phosphosulfate synthase	0.073491	0.080999	-0.028227	0.002335	-0.069363	0.091756	-0.091159	-0.080245	0.067129	0.049245	...	-0.042785	-0.081899	-0.041130	0.025566	0.122074	-0.021724	-0.085229	-0.029068	-0.036015	-0.100795
3-beta-hydroxysteroid dehydrogenase	0.067097	-0.061427	0.093204	0.108998	-0.041609	0.058034	0.041132	-0.040696	0.152901	-0.081870	...	-0.233139	-0.060815	0.187243	0.057241	-0.081594	0.062716	-0.078905	-0.121561	-0.014237	0.058866
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
pp10122	0.178080	0.371527	-0.412717	-0.246089	0.214775	-0.506345	-0.290790	-0.150410	0.215877	0.408559	...	-0.709345	-0.260173	0.380616	-0.316627	-0.490632	0.194519	0.108054	-0.426641	0.036487	-0.381604
tRNA-uridine aminocarboxypropyltransferase	-0.174120	0.032164	-0.087492	0.100593	-0.302876	0.200717	0.170120	-0.217411	-0.027259	0.027179	...	-0.102030	-0.308913	0.217483	-0.102495	-0.275602	-0.286097	0.077114	-0.144031	0.070092	-0.232707
tmp_locus_54	0.157015	0.218119	-1.155761	0.492320	0.409154	0.175829	-1.217417	-0.528736	-0.462023	0.198218	...	-0.510813	0.149230	0.147140	0.040833	-0.103283	-1.122915	-0.044513	-0.253034	-0.038325	-0.133388
urf-ret	0.335659	0.228930	0.175542	-0.229068	0.083526	0.178109	0.427678	0.007911	0.225716	0.223766	...	-0.747746	-0.156841	0.326407	0.113307	-0.329125	-0.432075	-0.565949	0.077020	0.458489	-0.419929
zf30	-0.410130	-0.274361	0.290211	-0.336239	0.221474	-0.332876	0.159841	-0.259432	0.078994	0.229157	...	-0.663979	-0.332803	0.177944	-0.310315	-0.063604	0.098105	0.360965	0.330712	0.027433	-0.002185

19334 rows × 128 columns