Skip to main content
Figure 1 | BMC Medical Genomics

Figure 1

From: Virtual CGH: an integrative approach to predict genetic abnormalities from gene expression microarray data applied in lymphoma

Figure 1

vCGH model structure. (A) vCGH model presented as a Bayesian network. The shaded nodes S 1 , S 2 , ..., S n represent hidden state variables for genes and the white nodes E 1 , E 2 , ..., E n represent the observations for the variables. There are three symbols for GEP observations, "H", "L" and "M" for high, low and medium expression, respectively. There are nine hidden states that GEP profiles superimposed on CGH, H + , L + , M + , H - , L - , M - , H o , L o and M o , where = "+", "-" and "o" for gain, loss and normal CGH status, respectively. (B) State transition diagram of vCGH model. The model is a single HMM chain integrating three Markov sub-chains: (+), (-) and (o). In each sub-chain, a Markov chain is graphically shown as a collection of states, with arrows between them describing the state transitions within a CNA (gain, loss or normal). There are also arrows between sub-chains, describing the state transitions from one CNA to another CNA.

Back to article page