Hamming approach degree is defined to the similar degree between the undetermined samples and standard samples, and the greater its value indicates that the undetermined samples are closer to the standard samples, that is to say, the undetermined samples and the standard samples are owned to the same class.
Figure 2b shows the curve of 27 undetermined samples and cluster center 7; and number 19, number 20, and number 21 undetermined samples' hamming approach degree are the biggest, 0.981, 0.979 and 0.979, respectively.
The classes of 27 undetermined samples were identified by analyzing the hamming approach degree curves.
In this paper, 27 undetermined samples are selected in the depth Section 100.0 5118.2 miles (the whole well).
The classes of the samples can be determined by analyzing the hamming approach degree curves, which is calculated by the undetermined samples and 9 cluster centers.
As shown in Fig. 2a, it is the hamming approach degree curve of 27 undetermined samples and cluster center 4 (Table 2); the hamming approach degrees of number 10, number 11, and number 12 undetermined samples are the biggest, 0.984, 0.985, and 0.976, respectively, almost 1.
The computation of the hamming approach degree is as follows (Zhang et al. 2010): H = 1 - sumlimits_{j = 1}^{n} {w_{j} left| {x_{j} - c_{k} } right|} ;quad k = 1,, 2, ldots,{text{ z}}, (9 where (H) was the hamming approach degree of the undetermined samples and one cluster center.
Undetermined samples were treated as missing data in all subsequent analyses.
Undetermined samples from an assay were sequenced if DNA was available, resulting in a final genotyping call rate of 99.5%, with call rates ≥ 98.8% for each SNP.
In the breast tumor samples, for example, the number of Undetermined samples decreases from 31too 4. Four of these 27 samples (14.8 %) were assigned to the correct tissue of origin, and 23 were not.
