protein-feature-vectors

Code to produce fixed-length feature vectors from protein sequences based on iFeatureOmega.

Introduction

This package is derived from iFeatureOmega-CLI, with the following differences:

• It only makes feature vectors using protein sequence (not based on DNA, RNA, or compounds)
• It only makes fixed-length feature vectors
• It has fewer dependencies than iFeatureOmega-CLI (i.e. just numpy, pandas, and fastapy)
• It just creates feature vectors and has no other plotting or analytical capabilities
• It has additional variants of specific algorithms (e.g. TPC_type_3, CKSAAP_type_3) created for faster Deep Learning

Installation

git clone git@github.com:bosborne/protein-feature-vectors.git
cd protein-feature-vectors
pip3 install .

Usage

Sequences from a Fasta file:

from protein_feature_vectors import Calculator
calc = Calculator()
calc.display_feature_types()
calc.get_feature_vectors("CTriad", file='data_examples/multi.fa')
print(calc.encodings)

Sequences from a dict:

from protein_feature_vectors import Calculator
calc = Calculator()
seqs = {"A1": "MLVTIKIQLKEAL...LDTGADVLEDM", "B2": "MHLPGKWMIGGIG....GFIKVRQYDEICGH"}
calc.get_feature_vectors("Moran", pdict=seqs)

Get ids and values from the encodings DataFrame as Series objects

calc.get_feature_vectors("CTDC", pdict=seqs)
protein_ids = [x[0] for x in calc.encodings.iterrows()]
values = [x[1].tolist() for x in calc.encodings.iterrows()]

Save data to a file

calc.get_feature_vectors("AAC", pdict=seqs)
calc.to_csv("AAC.csv", index=True, header=True)

Feature Vectors

Name	Description	Length
AAC	Amino acid composition	20
AC	Auto covariance	24
ACC	Auto-cross covariance	192
APAAC	Amphiphilic PAAC	26
ASDC	Adaptive skip dipeptide composition	400
CC	Cross covariance	168
CKSAAGP_type_1	Composition of k-spaced amino acid group pairs - normalized	100
CKSAAGP_type_2	Composition of k-spaced amino acid group pairs - raw count	100
CKSAAP_type_1	k-spaced amino acid pairs, k = 0,1,2,3 - normalized	1600
CKSAAP_type_2	k-spaced amino acid pairs, k = 0,1,2,3 - raw count	1600
CKSAAP_type_3	k-spaced amino acid pairs, k = 0,1,2,3 - normalized, rounded count	1600
CTDC	Composition	39
CTDD	Distribution	195
CTDT	Transition	39
CTriad	Triads of 7 groups based on dipoles and volumes	343
DDE	Composition and distribution of dipeptides	400
DistancePair	PseAAC of distance-pairs and reduced alphabet	20
DPC_type_1	Dipeptide composition - normalized	400
DPC_type_2	Dipeptide composition - raw count	400
GAAC	Grouped amino acid composition	5
GDPC_type_1	Grouped dipeptide composition - normalized	25
GDPC_type_2	Grouped dipeptide composition - raw count	25
Geary	Spatial autocorrelation of physicochemical properties	24
GTPC_type_1	Grouped tripeptide composition - normalized	125
GTPC_type_2	Grouped tripeptide composition - raw count	125
KSCTriad	k-spaced CTriad, k = 0,1,2,3	1372
Moran	Spatial autocorrelation of physicochemical properties	240
NMBroto	Normalized Moreau-Broto	240
PAAC	Pseudo-amino acid composition	23
PseKRAAC_type_1	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_2	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_3A	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_3B	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_4	Pseudo K-tuple reduced amino acids composition	25
PseKRAAC_type_5	Pseudo K-tuple reduced amino acids composition	9
PseKRAAC_type_6A	Pseudo K-tuple reduced amino acids composition	16
PseKRAAC_type_6B	Pseudo K-tuple reduced amino acids composition	25
PseKRAAC_type_6C	Pseudo K-tuple reduced amino acids composition	25
PseKRAAC_type_7	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_8	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_9	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_10	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_11	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_12	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_13	Pseudo K-tuple reduced amino acids composition	16
PseKRAAC_type_14	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_15	Pseudo K-tuple reduced amino acids composition	4
PseKRAAC_type_16	Pseudo K-tuple reduced amino acids composition	4
QSOrder	Quasi-sequence-order descriptors (nlag=13)	46
SOCNumber	Sequence-order-coupling number	6
TPC_type_1	Tripeptide composition - normalized	8000
TPC_type_2	Tripeptide composition - raw count	8000
TPC_type_3	Tripeptide composition - normalized, rounded count	8000