Safety in numbers: multiple occurrences of highly similar homologs among Azotobacter vinelandiicarbohydrate metabolism proteins probably confer adaptive benefits

Table 1 Summary of statistical data for SEED data sets ¹

	All synolog pairs			≥90% protein sequence identity between synolog pairs
	Median ± MAD	Min	Max	Median ± MAD	Min	Max
Number of carbohydrate metabolism synologs	49.0 ± 35.0	0	394	0.0 ± 0.0	0	47
Number of carbohydrate metabolism synolog groups	20.0 ± 13.0	0	128	0.0 ± 0.0	0	16
Average protein sequence identity between synolog pairs² [%]	36.8 ± 4.2	13.4	100.0	97.3 ± 1.8	90.0	100.0
Synolog fraction of carbohydrate metabolism genes [%]	30.0 ± 9.7	0.0	85.7	0.0 ± 0.0	0.0	34.6

¹Median, minimum and maximum values for the carbohydrate metabolism gene set extracted from 943 prokaryote genomes in the SEED database [19], with no set cutoff and with a cutoff set at 90% protein sequence identity between synologs. Synologs are here defined as intra-genome sequences assigned to the same FIGfam (see text). The synolog fraction describes the ratio of the total number of synologs relative to the total number of genes in a genome. MAD is median absolute deviation. The median number of carbohydrate metabolism genes in the data set was 160.0 ± 74.0. The minimum and maximum numbers of carbohydrate metabolism genes observed among the included genomes were 4 and 585 genes, respectively.
²Calculated from the genomes containing carbohydrate metabolism synologs at the given cutoff.

ISSN: 1471-2164