Genetics. In the article “Identification of protein coding regions in the human genome by quadrant discriminant analysis,” by M. Q. Zhang, which appeared in number 2, January 23, 1997, of Proc. Natl. Acad. Sci. USA (94, 565–568), the author wishes to publish a correction to Table 2. The second entry in column one should read HEXON instead of FGENEH. A complete corrected table appears below:

Table 2.

Test summary

Program	sn	sp	SN	SP	CC
GRAIL2	0.51	0.57	0.79	0.85	0.80
HEXON	0.71	0.65	0.88	0.80	0.83
MZEF	0.84	0.92	0.88	0.95	0.90

See Methods for definitions. Lowercase letters indicate the exon level; uppercase, the base-pair level.