Interpretation of microbiota-based diagnostics by explaining individual classifier decisions

BMC Bioinformatics

Table 3 Most relevant bacterial species for the top IBD microbiota sample presented in Fig. 3

Species name	IS-pro peak(s)	Presence	Relevance
Escherichia coli	735	+	0.62
Alistipes finegoldii	396	–	0.57
Alistipes finegoldii	230	–	0.52
unclassified Proteobacterium	747	+	0.35
Lachnospiraceae sp.	606	+	0.3
Alistipes putredinis	235	–	0.29
unclassified Firmicute	326	+	−0.26
Finegoldia magna	306	+	0.24
Parabacteroides distasonis	461	+	0.23
Morganella morganii	855	+	−0.23
Alistipes finegoldii	400	–	0.21
Bacteroides fragilis	605	+	−0.21
Escherichia coli	827	+	0.19
Alistipes putredinis	236	–	0.18
Streptococcus sp.	321	+	0.17
unclassified Firmicute	525	+	0.17
Bacteroides vulgatus	479	–	0.16
Bacteroides sp.	594	+	0.16
Alistipes finegoldii	407	–	0.15
Escherichia coli	828	–	−0.14
Bacteroides fragilis	537	–	0.05

’+’ and’-’ indicate whether a peak is present or absent in a sample, and needs to be coupled with the relevance for interpretation. Ranking is based on absolute relevance values; Classification was done by a linear SVM

ISSN: 1471-2105