Mass spectrometry protein expression profiles in colorectal cancer tissue associated with clinico-pathological features of disease

Table 3 Performance of predictive models for classification of clinico-pathological characteristics in normal mucosa tissue.

CHARACTERISTICS	¹Advanced Dukes' stage	Poorly differentiated	Lymph node involvement	Invasiveness	²Disease recurrence
Number of features	7	5	3	6	7
Positive prediction rate	8/13	8/14	11/14	3/7	0/5
Sensitivity	0.615	0.571	0.786	0.429	0.000
³CI	0.322-0.849	0.296-0.812	0.488-0.943	0.116-0.798	0.000-0.537
Positive predictive value	0.500	0.444	0.733	0.500	0.000
CI	0.255-0.749	0.224-0.686	0.448-0.911	0.139-0.860	0.000-0.945
Negative prediction rate	12/20	9/19	15/19	23/26	25/26
Specificity	0.600	0.474	0.789	0.885	0.962
CI	0.364-0.800	0.252-0.705	0.539-0.930	0.687-0.970	0.784-0.998
Negative predictive value	0.706	0.600	0.833	0.852	0.833
CI	0.440-0.886	0.329-0.825	0.577-0.956	0.654-0.951	0.645-0.937
Absolute error	0.394	0.485	0.212	0.212	0.194
⁴ROC error	0.392	0.477	0.212	0.343	0.519
Fisher's exact test	P = 0.139	P = 0.267	P = 0.001	P = 0.082	P = 0.839

¹Includes Dukes' C1 and C2; ² Median follow-up time for recurrent disease patients: 33 months; median follow-up time for disease-free patient: 27 months (analysis excludes patients who died through surgical complications - see Table 1); ³CI = 95% confidence interval; ⁴ROC = receiver-operator characteristics
The KNN algorithm [29] was used in 'leave-one-out' cross-validation prediction with the number of features (marker peaks) specified. Marker peaks were selected using a t-test statistic except for lymph node involvement and invasiveness characteristics of tumour tissue where the SNR test statistic was used.

ISSN: 1471-2407