Using Python CLI

For local processing or automation, use the ProtSpace Python package.

Installation

pip install protspace

Commands

From UniProt Query

protspace-query -q "(ft_domain:kinase) AND (reviewed:true)" -m pca2,umap2

From Your Own Embeddings

protspace-local -i embeddings.h5 -m pca2,umap2

Parameters

Both commands share similar parameters. The key difference:

• protspace-query uses -q to specify a UniProt query
• protspace-local uses -i to specify an input embeddings file

Parameter	Description	Command
-q	UniProt query string	query only
-i	Input embeddings (HDF5)	local only
-o	Output directory	both
-m	Projection methods	both
-f	Annotations (names or CSV path)	both

Annotations

Specify annotations with -f:

# By name (auto-retrieved)
-f protein_family,reviewed,pfam,genus,species

# Or provide a CSV file
-f annotations.csv

CSV format:

identifier,taxonomy,family,function
P12345,Bacteria,Kinase,ATP binding
P67890,Archaea,Phosphatase,Hydrolase
Q54321,Eukaryota,Kinase,Transferase

The identifier column must match protein IDs in your embeddings file.

Projection Methods

Methods require a dimension suffix: 2 for 2D, 3 for 3D.

Dimension Suffix Required

Specify pca2 or pca3, not pca alone - the dimension suffix is mandatory.

Method	2D	3D	Description
PCA	pca2	pca3	Principal Component Analysis
UMAP	umap2	umap3	Uniform Manifold Approximation
t-SNE	tsne2	tsne3	t-distributed Stochastic Neighbor Emb.
PaCMAP	pacmap2	pacmap3	Pairwise Controlled Manifold Approx.
MDS	mds2	mds3	Multidimensional Scaling

TIP

ProtSpace is optimized for 2D visualization - prefer *2 methods over *3.

More Info

Find full docs and more examples on the ProtSpace Python GitHub.