Table 3.

Characterizing the 20 most anomalous 1 kb segments of the E. coli K12 genome based on dinucleotide signature dissimilarity as measured by (a) DRA-divergence, (b) delta-distance, (c) Euclidean distance and (d) the quadratic discriminant

#	(a) chi-square		(b) delta-distance		(c) Euclidean		(d) quadratic		(e) G + C
	loc.	HT	loc.	HT	loc.	HT	loc.	HT	loc.	HT
1	200	yaeT	151	−1	151	−1	284	M	583	M
2	227	0	227	0	274	M	525	M	584	M
3	393	M	394	−1	575	M	526	M	1212	−1, −1
4	394	−1	526	M	777	−1	575	M	1636	M
5	526	M	777	−1	978	mukB	777	−1	2102	−1
6	777	−1	1287	−1	1142	rne	1142	rne	2105	+1
7	978	mukB	1427	M	1395	M	1427	M	2468	M
8	1142	rne	1465	0	1427	M	1465	0	2773	M
9	1465	0	1527	0	1465	0	1527	0	2783	0
10	1527	0	1707	−1	1527	0	1707	−1	2785	0
11	1707	−1	2071	M	2101	M	2104	+1	2989	+1
12	2071	M	2072	M	2104	+1	2105	+1	2990	−1
13	2072	M	2104	+1	2105	+1	2989	+1	2994	0
14	2104	+1	2994	0	2994	0	2990	−1	3267	+1, +1
15	3312	−1, infB	3312	−1, infB	3314	−1	2992	−1	3581	+1
16	3450	rplWD	3602	ftsY	3450	rplWD	2994	0	3797	+1,+1
17	3602	ftsY	3915	−1	3602	ftsY	3602	ftsY	3798	−1
18	3915	−1	4058	−1	4121	−1	3620	M	3803	−1, +1
19	4058	−1	4187	rpoC	4503	M	4503	M	4267	−1
20	4181	rpoB	4474	−1, +1	4504	M	4504	M, M	4475	+1
Genes	19		18		18		18		21
Errors	14(8)		12(3)		9(5)		6(2)		7(0)

Each segment, indexed by location (loc.) in kb from the published origin, overlaps zero, one or two genes in the protein table. Intergenic segments are classified as ‘0’. HT is indicated by ‘+1’ or by ‘M’ (if the gene is a mobile element). False positives are indicated by gene locus (if essential) or by ‘−1’ (if not). False positives account for total errors (bottom row) and the number of essential genes is given (in parentheses). The analysis is repeated for the 20 most anomalous segments with respect to (e) GC-divergence.