Predicting the pathogenicity of bacterial genomes using widely spread protein families

BMC Bioinformatics

Table 1 The top HP PGFams that serve as features of WSPC according to their average MDI rank, along with the number of HP and NHP genomes in the training set that contain the respective PGFams

	PGFam ID	Gene function	General function	MDI (SD\(^1\))	HPs	NHPs	P-ratio\(^2\)	# Genera\(^3\)
1	04139053	Uroporphyrinogen III decarboxylase	Energy production	0.038 (0.012)	362	27	6.47	109
2	01915472	Dihydrolipoamide acetyltransferase component of PDC\(^{*}\)	Aerobic respiration	0.035 (0.01)	385	48	3.93	120
3	07629184	Cytosol aminopeptidase PepA	Regulation	0.03 (0.009)	366	39	4.58	115
4	07157721	Heme O synthase, protoheme IX farnesyltransferase, COX10-CtaB	Aerobic respiration	0.022 (0.007)	312	14	10.41	89
5	00022550	Molybdopterin synthase catalytic subunit MoaE	Respiration, energy conversion	0.013 (0.005)	303	17	8.42	89
6	01033770	Class 2 Dihydroorotate dehydrogenase (DHODase)	Amino acid biosynthesis	0.011 (0.005)	333	35	4.63	99
7	00006100	tRNA-modifying protein YgfZ	Synthesis and repair	0.011 (0.006)	305	17	8.48	93
8	07941512	23S rRNA methyltransferase	Methylation	0.01 (0.003)	324	37	4.27	89
9	00405499	YpfJ protein, zinc metalloprotease superfamily	Protein cleavage	0.009 (0.003)	273	13	9.76	87
10	06757295	Threonine dehydratase	Amino acid biosynthesis	0.008 (0.004)	323	34	4.62	95
11	08199696	Glutathione reductase	Stress tolerance	0.008 (0.002)	220	12	8.48	66
12	03081665	Cell division integral membrane protein, YggT	Stress tolerance	0.007 (0.003)	352	56	3.09	109
13	07854425	Superoxide dismutase [Cu–Zn] precursor	Stress tolerance	0.006 (0.001)	252	21	5.74	75
14	01668012	Sulfur carrier protein FdhD	Stress tolerance	0.006 (0.003)	300	26	5.56	98
15	01147190	Deoxyribodipyrimidine photolyase	DNA repair	0.006 (0.003)	281	17	7.82	88

The average MDI rank of a PGFam is the average value of the feature’s MDI values computed using 100 random forest classifiers with different random seeds trained on the training set. \(^1\)Standard Deviation. \(^2\)The ratio between the proportion of HPs with the corresponding PGFam and the proportion of NHPs with the corresponding PGFam. To avoid zero division, add-one smoothing was performed. \(^3\) The number of different genera to which the genomes that contain the respective PGFams belong. PDC: Pyruvate dehydrogenase complex

ISSN: 1471-2105