kprot

kprot is a Python package for protein related analysis.

The sections below introduce the package and show quick-start examples from each exported module.

Installation

pip install kprot

Quick start

The examples below follow the notebooks under nbs/ in order. Each function example lives in its own cell and starts with a short comment derived from the function docstring.

01 utils

from kprot.uniprot import get_uniprot_seq, get_uniprot_features, get_uniprot_kd, get_uniprot_type, apply_mut_single, apply_mut_complex, compare_seq
import re
from functools import lru_cache
import requests
from Bio.Align import PairwiseAligner
# Queries the UniProt database to retrieve the protein sequence for a given UniProt ID.
get_uniprot_seq("P04626")[:60]
'MELAALCRWGLLLALLPPGAASTQVCTGTDMKLRLPASPETHLDMLRHLYQGCQVVQGNL'
# Given a UniProt ID, fetch the protein metadata and annotated features.
get_uniprot_features("P04626").keys()
dict_keys(['uniprot_id', 'protein_name', 'gene_name', 'features'])
# Query Domain: Protein kinase entries based on a UniProt ID.
get_uniprot_kd("P04626")
[{'uniprot_id': 'P04626',
  'protein_name': 'Receptor tyrosine-protein kinase erbB-2',
  'gene_name': 'ERBB2',
  'start': 720,
  'end': 987,
  'description': 'Protein kinase',
  'sequence': 'LRKVKVLGSGAFGTVYKGIWIPDGENVKIPVAIKVLRENTSPKANKEILDEAYVMAGVGSPYVSRLLGICLTSTVQLVTQLMPYGCLLDHVRENRGRLGSQDLLNWCMQIAKGMSYLEDVRLVHRDLAARNVLVKSPNHVKITDFGLARLLDIDETEYHADGGKVPIKWMALESILRRRFTHQSDVWSYGVTVWELMTFGAKPYDGIPAREIPDLLEKGERLPQPPICTIDVYMIMVKCWMIDSECRPRFRELVSEFSRMARDPQRFV'}]
# Get region sequences for a UniProt feature type.
get_uniprot_type("P04626", "Signal")[:1]
[{'uniprot_id': 'P04626',
  'type': 'Signal',
  'protein_name': 'Receptor tyrosine-protein kinase erbB-2',
  'gene_name': 'ERBB2',
  'start': 1,
  'end': 22,
  'description': '',
  'sequence': 'MELAALCRWGLLLALLPPGAAS'}]
# Apply point mutations to a protein sequence.
apply_mut_single("MEEPQ", "M1A", "E2S")
Converted: M1A
Converted: E2S
'ASEPQ'
# Apply a composite mutation string to a protein sequence.
her2_seq = "LRKVKVLGSGAFGTVYKGIWIPDGENVKIPVAIKVLRENTSPKANKEILDEAYVMAGVGSPYVSRLLGICLTSTVQLVTQLMPYGCLLDHVRENRGRLGSQDLLNWCMQIAKGMSYLEDVRLVHRDLAARNVLVKSPNHVKITDFGLARLLDIDETEYHADGGKVPIKWMALESILRRRFTHQSDVWSYGVTVWELMTFGAKPYDGIPAREIPDLLEKGERLPQPPICTIDVYMIMVKCWMIDSECRPRFRELVSEFSRMARDPQRFV"
apply_mut_complex(her2_seq, "G776delinsVC/S783C", start_pos=720)[:25]
'LRKVKVLGSGAFGTVYKGIWIPDGE'
# Align two protein sequences and summarize the differences.
print(compare_seq("MEEPQ", "MAEPQ", return_text=True))
Original       1-5    : MEEPQ
Mutant                : MAEPQ
                         ^   

Differences:
  substitution at    2: E → A

02 onehot kmeans

from kprot.onehot import onehot_encode, onehot_encode_df, run_kmeans, filter_range_columns, get_clusters_elbow
import pandas as pd

# Set up the objects used by the examples below.
df = pd.DataFrame(
    {
        "site_seq": [
            "AASTYGVAKR",
            "GGSTYGVAKR",
            "AASTFGVAKR",
            "PPSTYGVAKK",
            "VVSTYGIARR",
        ]
    }
)
df.shape

df.head()
site_seq
0 AASTYGVAKR
1 GGSTYGVAKR
2 AASTFGVAKR
3 PPSTYGVAKK
4 VVSTYGIARR 
# One-hot encode aligned protein sequence windows.
onehot_encode(df["site_seq"]).head()
-20A -20G -20P -20V -19A -19G -19P -19V -18S -17T -16F -16Y -15G -14I -14V -13A -12K -12R -11K -11R
0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 1.0 0.0 1.0 1.0 0.0 1.0 1.0 1.0 0.0 0.0 1.0
1 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 1.0 1.0 0.0 1.0 1.0 0.0 1.0 1.0 1.0 0.0 0.0 1.0
2 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 1.0 1.0 0.0 1.0 0.0 1.0 1.0 1.0 0.0 0.0 1.0
3 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 1.0 1.0 0.0 1.0 1.0 0.0 1.0 1.0 1.0 0.0 1.0 0.0
4 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 1.0 1.0 0.0 1.0 1.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 
# One-hot encode a sequence column from a dataframe.
onehot = onehot_encode_df(df, seq_col="site_seq")
onehot.head()
-20A -20G -20P -20V -19A -19G -19P -19V -18S -17T -16F -16Y -15G -14I -14V -13A -12K -12R -11K -11R
0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 1.0 0.0 1.0 1.0 0.0 1.0 1.0 1.0 0.0 0.0 1.0
1 0.0 1.0 0.0 0.0 0.0 1.0 0.0 0.0 1.0 1.0 0.0 1.0 1.0 0.0 1.0 1.0 1.0 0.0 0.0 1.0
2 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 1.0 1.0 0.0 1.0 0.0 1.0 1.0 1.0 0.0 0.0 1.0
3 0.0 0.0 1.0 0.0 0.0 0.0 1.0 0.0 1.0 1.0 0.0 1.0 1.0 0.0 1.0 1.0 1.0 0.0 1.0 0.0
4 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 1.0 1.0 0.0 1.0 1.0 1.0 0.0 1.0 0.0 1.0 0.0 1.0 
# Fit KMeans to one-hot encoded features and return the assigned labels.
run_kmeans(onehot, n=2)
array([0, 0, 0, 0, 1], dtype=int32)
# Filter one-hot columns to a position window around the center residue.
onehot_10 = filter_range_columns(onehot, low=-10, high=10)
onehot_10.head()
0
1
2
3
4 
# Plot the elbow curve for a range of KMeans cluster counts.
get_clusters_elbow(onehot, max_cluster=5, interval=2)

03 embeddings

from kprot.embeddings import get_esm, get_t5, get_t5_bfd
import pandas as pd

# Set up the objects used by the examples below.
df = pd.DataFrame(
    {
        "sequence": [
            "MKWVTFISLLLLFSSAYSRGVFRRDAHKSEVAHRFKDLGE",
            "GSSHHHHHHSSGLVPRGSHMNNIRRVAILAALVAGQALA",
        ]
    }
)
df.shape

df.head()

run_model_examples = False
# Run the example.
df = pd.DataFrame(
    {
        "sequence": [
            "MKWVTFISLLLLFSSAYSRGVFRRDAHKSEVAHRFKDLGE",
            "GSSHHHHHHSSGLVPRGSHMNNIRRVAILAALVAGQALA",
        ]
    }
)
df.shape
(2, 1)
# Extract ESM2 embeddings (mean pooled per sequence).
if run_model_examples:
    get_esm(df, "sequence").head()
else:
    print("Skipped get_esm example: set run_model_examples=True to allow model downloads.")
Skipped get_esm example: set run_model_examples=True to allow model downloads.
# Extract ProtT5-XL-uniref50 embeddings from protein sequences in a dataframe.
if run_model_examples:
    get_t5(df, "sequence").head()
else:
    print("Skipped get_t5 example: set run_model_examples=True to allow model downloads.")
Skipped get_t5 example: set run_model_examples=True to allow model downloads.
# Extract ProtT5-XL-BFD embeddings from protein sequences in a dataframe.
if run_model_examples:
    get_t5_bfd(df, "sequence").head()
else:
    print("Skipped get_t5_bfd example: set run_model_examples=True to allow model downloads.")
Skipped get_t5_bfd example: set run_model_examples=True to allow model downloads.