Python Programming for Biology: Bioinformatics and Beyond

Figure 19.5 (Plate 9).  Example of 2D peak picking for a protein gel. An image of a

two-dimensional polyacrylamide gel, with the peak picked maxima shown as small

crosses. The gel has been stained with Coomassie Brilliant Blue dye to show the

abundance of different proteins that have been separated according their size and

isoelectric value.

By looping though the resulting Peak objects we can extract the identified positions and

use  this  to  mark  small  crosses  in  the  image  pixmap  that  we  initially  loaded,  here  setting

the RGB colour for the pixel to be yellow, maximum red and green but no blue. Note the

use  of  the  array  slice  notation  [xIndices,  yIndices,:],  which  sets  the  whole  of  the  last

(colour)  axis  for  the  required  rows  and  columns  (which  are  specified  in  the  tuples  of

indices).  The  pixmap  is  then  converted  into  a  PIL  Image  object  so  that  it  can  easily  be

viewed and saved to file.

color = (255.0, 255.0, 0.0)

for peak in peaks:

x,y = peak.position

xIndices = (x-1, x, x, x, x+1) # X indices for a cross shape

yIndices = (y, y-1, y, y+1, y) # Y indices for a cross shape

pixmap[xIndices, yIndices,:] = color

img2 = pixmapToImage(pixmap, mode='RGB')

img2.show()

img2.save('PickedGel.png')

If  we  wish  to  refine  the  peak  locations,  i.e.  to  interpolate  their  centres  between  the

integer grid points, it is a simple matter of calling the .fit() method of the peak and looking

at the the updated attribute values.

for peak in peaks:

peak.fit(fitWidth=3)

print(peak.fitAmplitude, peak.fitAmplitude, peak.fitLinewidth)

---