Bioinformatic Analysis of Genetic Factors from Human Blood Samples and Postmortem Brains in Parkinson’s Disease

Figure 2

Consensus clustering provides quantitative evidence for determining the number and membership of possible clusters in the datasets. (a) Volcano plots. (b) CDF curve. (c, d) Consensus clustering delta area curve (c) and column chart (d) displaying the relative change in area under the cumulative distribution function (CDF) curve for each category number compared to –1. The vertical axis depicts the relative change in area under the CDF curve, while the horizontal axis reflects the category number k. (e, f, d) Consensus sample cluster heat map. Different samples are represented by the rows and columns of the matrix. The values of the consensus matrix range from 0 (cannot be clustered together) to 1 (always clustered together) in yellow to dark red.

(a)

(b)

(c)

(d)

(e)

(f)