Page 60 - TD-4-1

Tumor Discovery Drug repurposing for pancreatic cancer via AI

2.4. Elimination of false positives from candidate for x = 1,2…,X-1,X, n = 1,2…,N-1,N
genome-wide genetic and epigenetic networks for
real genome-wide genetic and epigenetic networks    
of PDAC and healthy controls based on system  
identification and order detection methods  
 
Based on the stochastic system models of protein interaction  
and gene regulations described above, we constructed four  
interaction and regulation models for candidate GWGENs,  
including the candidate PPI model, candidate gene regulation    
model, candidate lncRNA regulation model, and candidate  
miRNA regulation model in Equations I–IV. Next, we pruned  t n   1  t  1  l   ζ 
nl n  
false positive interactions and regulations, from these candidate  y S  U y    y− 1  

nm nl n  

GWGENs using system identification and order detection l n =  y   1  y   ×   
methods, utilizing microarray data from both PDAC and non-   m   1    ζ yS  −
nl n
y  
PDAC samples to estimate the corresponding real GWGENs.  V y   y 
 ψ y− 1 
Equations I–IV can be reformulated as the following   
linear regression equations to obtain parameter vectors for  ψ 
interactions and gene regulations.   yU− y  
n
p n p 1 w 2 p R w   − κ y− 1  
n p 1
n pn p
n pn
w
w

w

 
 − κ yV−  y
 β 
 y 
w1 δ + n =  φ n θ  + δ ⋅ n 
n
n

w2 w PPIN w w PPIN (V) y  y  y y   (VII)
n

w

for y =1,2…,Y-1,Y, n = 1,2…,N-1,N

wR w  
w PPIN  
 
for w =1,2…,W-1,W, n = 1,2…,N-1,N  
   
   
   
   
   
   
   t n   t    nl n l   
   1  z S  1  U z   λ z− 1 
  m n =     ×   

nm nm n  

  z   1  z    
nm n 
   m  z  1     λ zS 
z V 
−
 α   µ z 
 x− 1   z− 1 
     
 t n   t     α   
nl n
1 
1 
 x S    xS x   µ z U z 
−
−
n

n m n g  
n = 
g  l  1  x   ×  γ x− 1   − ρ 1  
x 
x 
U
      z−
ng 
n
x 

  m  x  1      
V
 γ xU   − ρ 
−
 x   zV z 
−
 ε − x− 1    β z  
  
n = 
n 
n θ 
  δ + z  φ z  z + δ ⋅ z  (VIII)

 ε − xV x 
−
 β 
n θ 
n = 
δ + x  φ x  x + δ ⋅ x   x  (VI) for z =1,2…,Z-1,Z, n = 1,2…,N-1,N
n 
Volume 4 Issue 1 (2025) 52 doi: 10.36922/td.4709