Reads were again generated, and treated as the dataset for the sample 2. To test the effects of sequencing depth, we varied the number of generated reads for each dataset from 1 to 50 million (M).
These reads were treated as the dataset for the sample 1. Next, 100 random regions were defined as DMRs for Up or Down, and methylation levels of all CpG sites in these regions were changed to the maximum or the minimum, respectively.
The datasets for the sampled individuals can be assigned to taxa according to the results of the analysis and the character data can be aggregated to add to the taxon characterization.
As shown in Fig. 3A, CCScycling and CCSphase scores for the total gene dataset for the normal samples were consistently very low, while scores for tumors were varying degrees higher, indicating variation in the proportion of cycling cells.
Larger protein complexes generally display increased diversity in terms of interaction types (except for complexes of size 8 in the HeLa dataset for which the sample size is very small), probably because they may contain functionally specialized subcomplexes, each with its own prevailing interaction type.
For Dataset 2, the samples on Chip12 had lower average β values than those in Chip11 and shifted to the left in the density plot.
We compared the SNVs in each pool to the SNVs reported in the HM+1KG dataset for the corresponding 20 samples in the pool.
To compile quantitative measurement of the identified peptides, the datasets for the two samples were processed for finding features (unique mass classes) based on the algorithm described previously [41].
Samples were excluded from the dataset for the reasons of sample failure, sex mismatch and first-degree relative of an included individual based on genotype data.
HW, HY, JA, GH were involved in collecting the samples in the dataset, processing the samples for the SELDI analysis, SELDI MS acquisition, data processing with CiphergenExpress and in verifying the quality of the peaks (section 2.3).
