Laboratory methods to improve SELDI peak detection and quantitation

Table 2 Summary table of results comparing Exp.1 and Exp.2 showing improvement in peak detection and peak intensity variation following optimization of protocols.

		Experiment 1					Experiment 2
Protein Chip	Fraction	No. Spectra excluded (Total 18)	No. Peaks	Averaged CV Peak Intensity (%)	CV Range	No. Peaks (%) statistically different across batches (p < 0.01)^a	No. Spectra excluded (Total 18)	No. Peaks	Averaged CV Peak Intensity (%)	CV Range	No. Peaks statistically (%) different across batches (p < 0.01)^a
	F1	5	16	31.7	12–127	0	1	28	20.9	8–36	2 (7)
	F3	1	11	31.2	15–75	0	0	21	31.4	12–52	2 (10)
IMAC	F4	3	7	44.7	29–61	1 (14)	0	20	26.0	14–45	2 (10)
	F6	3	21	27.0	15–39	0	7	27	22.6	11–49	9 (33)
	LS-F1	4	22	31.0	15–50	0	1	23	18.1	11–41	1 (4)
	LS-F3	6	5	41.5	25–48	0	0	23	26.4	13–40	3 (13)
	LS-F4	1	13	37.4	18–55	6 (46)	1	26	14.7	8–30	0
CM10	LS-F5	4	9	30.6	20–46	3 (33)	0	27	16.4	8–30	1 (4)
	LS-F6	3	13	29.4	18–44	0	2	30	25.4	8–50	11 (37)
	HS-F3	6	3	66.5	53–85	0^b	1	15	49.2	27–148	9 (60)
	HS-F4	6	5	48.7	19–72	0	3	14	30.6	10–46	0
	F1	5	6	67.5	40–129	0	0	16	32.7	15–66	2 (13)
	F3	0	13	32.4	16–50	3 (23)	1	17	22.8	15–38	5 (29)
H50	F4	2	15	33.0	15–79	2 (15)	1	21	22.5	11–53	5 (24)
	F6	2	30	23.7	13–39	1 (3)	1	35	25.0	12–55	3 (9)

^a This column reports the number of peaks in the ProteinChip-fraction spectra that were considered to have different peak intensities between batches as calculated by the non-parametric Kruskal-Wallis (p < 0.01), the data was bootstrapped using 2000 randomizations of the experiments to correct for multiple testing.
^b Analysis performed on 2 batches, not 3.

ISSN: 1477-5956