Examplary walkthrough analysis

Set paths

Ensure the current folder is set to QuimP11_MATLAB, or add the QuimP11_MATLAB to to [File->Set Path] so QuimP's functions are useable.

addpath('QuimP11_MATLAB')

Read in the data output by QuimP.

Missing files will be skipped. When prompted, select the folder containing the associated paQP file.

qCells = readQanalysis('example_output_new');
Read Q Analysis
Searching in "example_output_new/"
	...and its sub directories
Found 1 parameter files
Reading analysis: QW_channel_2_seg_0.paQP
	Warning: Map file could not be opened:
	example_output_new/QW_channel_2_seg_0_fluoCh3.maQP

qCells is a structure. Each element is data for one analysis (i.e. one paQP file). If only one paQP file was located then qCells will be of length of 1.

% we will extract the first analysis into 'c'.
c = qCells(1);
c  % type c and you will see a list of the data in the analysis
c = struct with fields:
               name: 'QW_channel_2_seg_0'
              index: 1
               PATH: 'example_output_new/'
          PARAMFILE: 'QW_channel_2_seg_0.paQP'
          SNAKEFILE: 'QW_channel_2_seg_0.snQP'
          STATSFILE: 'QW_channel_2_seg_0.stQP.csv'
    MOTILITYMAPFILE: 'QW_channel_2_seg_0_motilityMap.maQP'
     FLUOCH1MAPFILE: 'QW_channel_2_seg_0_fluoCh1.maQP'
     FLUOCH2MAPFILE: 'QW_channel_2_seg_0_fluoCh2.maQP'
     FLUOCH3MAPFILE: 'QW_channel_2_seg_0_fluoCh3.maQP'
      CONVEXMAPFILE: 'QW_channel_2_seg_0_convexityMap.maQP'
       COORDMAPFILE: 'QW_channel_2_seg_0_coordMap.maQP'
      ORIGINMAPFILE: 'QW_channel_2_seg_0_originMap.maQP'
           XMAPFILE: 'QW_channel_2_seg_0_xMap.maQP'
           YMAPFILE: 'QW_channel_2_seg_0_yMap.maQP'
           outlines: {40×1 cell}
               fluo: {40×1 cell}
                  R: [3 480 18 254]
           maxSpeed: [1×40 double]
     outlineHeaders: {'1.Position'  '2.X-coord'  '3.Y-coord'  '4.Origin'  '5.Global origin'  '6.Speed'}
        fluoHeaders: {'1.Intensity'  '2.X-coord'  '3.Y-coord'}
           nbFrames: 40
              stats: [40×11 double]
          fluoStats: [40×3×11 double]
        statHeaders: {'1.Frame'  '2.x-Centroid'  '3.Y-Centroid'  '4.Displacement'  '5.Distance Traveled'  '6.Directionality'  '7.Speed'  '8.Perimeter'  '9.Elongation'  '10.Circularity'  '11.Area'}
    fluoStatHeaders: {'1.Frame'  '2.Total Fluo.'  '3.Mean Fluo.'  '4.Cortex Width'  '5.Cyto. Area'  '6.Total Cyto. Fluo.'  '7.Mean Cyto. Fluo.'  '8.Cortex Area'  '9.Total Cortex Fluo.'  '10.Mean Cortex Fluo.'  '11.%age Cortex Fluo.'}
            SEGTIFF: '/home/baniuk/Documents/NEUBIAS/OSSL/QuimP_full-User_Manual/QW_channel_2_seg.tif'
                 PS: 0.2000
                 FI: 2
        cortexWidth: [1.5000 1.5000 -1]
         startFrame: 1
           endFrame: 40
        FLUOCH1TIFF: '/home/baniuk/Documents/NEUBIAS/OSSL/QuimP_full-User_Manual/QW_channel_1_actin.tif'
        FLUOCH2TIFF: '/home/baniuk/Documents/NEUBIAS/OSSL/QuimP_full-User_Manual/QW_channel_2_neg.tif'
        FLUOCH3TIFF: '/'
             frames: [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40]
        motilityMap: [40×400 double]
         fluoCh1Map: [40×400 double]
         fluoCh2Map: [40×400 double]
         fluoCh3Map: []
           coordMap: [40×400 double]
          originMap: [40×400 double]
          convexMap: [40×400 double]
               xMap: [40×400 double]
               yMap: [40×400 double]
         forwardMap: [40×400 double]
        backwardMap: [40×400 double]
             disMap: []

to access data, for example the fluoMap, type c.fluoMap

Load the image used for channel 1

If this fails, the path c.FLUOCH1TIFF is probably wrong due to it being moved

info = imfinfo(c.FLUOCH1TIFF,'TIFF'); % structure array with data about each image in the stack

im = imread(c.FLUOCH1TIFF, 3, 'Info', info); % load frame 3

figure(1);

imagesc(im);

axis equal

axis off

title('3rd frame of QW\_channel\_1\_actin.tif')

Plot the motility and fluo maps

figure(2);

plotMap(c.motilityMap, 'm', c.FI); % 'm' tells plotMap to format as a motility map.

% c.FI is the frame interval for scaling.

figure(3);

plotMap(c.fluoCh1Map, 'f', c.FI); % 'f' for fluorescence map

Calculate a cross correlation

xFM = xcorrQ( c.motilityMap,c.fluoCh1Map );

figure(4);

imagesc(xFM); title('Motility Map Xcorr with Fluo Map');

colorbar

yticks(0:5:40);

yticklabels(yticks*c.FI);

ylabel('Time (seconds)')

xticks(0:100:400);

xticklabels(xticks./400);

xlabel('Cell outline')

Yellow is strong correlation, blue strong negative correlation. We can also do auto-correlation ( correlate maps with themselves) to search for repetitive patterns of motility/fluorescence.

xMM = xcorrQ( c.motilityMap, c.motilityMap); % calc an auto correlation

figure(5);

imagesc(xMM); title('AutoCorr of Motility Map');

colorbar

yticks(0:5:40);

yticklabels(yticks*c.FI);

ylabel('Time (seconds)')

xticks(0:100:400);

xticklabels(xticks./400);

xlabel('Cell outline')

Plot the cell outlines, stored in c.outlines.

This a cell array. Each cell as the outline for one frame. See c.outlineHeaders for the contents of the columns. We want columns 3 and 4.

c.outlineHeaders

c.outlines{7}(:,3:4) % x and y co-ordinates at frame 7

figure(6);

colours = summer(c.nbFrames); % create a colour chart

hold off;

% plot all the cell outline

for i = 1:c.nbFrames

plotOutline(c.outlines{i}, colours(i,:));

hold on;

end

hold off;

axis equal;

axis(c.R); % set the axis bounds

set(gca,'YDir','reverse'); % flip the Y-axis around so matlab plots (0,0) at the top left

Plot and colour and outline according to membrane speed

The maxMigration value is used to scale the colours

figure(7);

% we take the max velocity across all frames and include negative values

scale = max(abs(c.maxSpeed));

% frame 35. red is expansion, blue contraction

plotMotility(c.outlines{35}, scale, 5);

axis(c.R); % set the axis bounds

axis equal;

% flip the Y-axis around so matlab plots (0,0) at the top left

set(gca,'YDir','reverse');

% this function needs a bit of work to improve its visuals

Plot some global statistics

c.stats is a matrix, rows are frames. See c.statHeaders for column contents.

The cell centroid is computed as the weighted centre of the polygon formed by the cell outline. The measures computed by BOA are as follows (http://pilip.lnx.warwick.ac.uk/docs/master/QuimP_Guide.html):

X-Centroid [pixels] - Horizontal pixel co-ordinate of the cell centroid.
Y-Centroid [pixels] - Vertical pixel co-ordinate of the cell centroid.
Displacement [microns] - Distance the cell centroid has moved from its position in the first recorded frame.
Distance Travelled [microns] - Sum of the centroid displacements between frame, i.e. the total distance over which the centroid has moved.
Directionality - Persistence in direction, calculated as Displacement∕Dist.Travelled (chemotaxis index). A value of 1 reveals that a cell has moved in a straight line. Decreasing values denote a cell moving increasingly erratically.
Speed [microns per second] - Speed at which the centroid moved between the current and previous frame.
Perimeter [microns] - Length of the cell perimeter (segmented outline).
Elongation - An ellipse is fitted to the cell outline and the major/minor axis used to compute the elongation of the cell’s shape. Elongation = major axis∕minor axis. Note that a value of 1 does not necessarily represent a perfectly circular cell, only a circular fitted ellipse.
Circularity - A measure of circularity defined by the following equation: . A value of 1 reveals the cell’s outline to be perfectly circular.
Area [microns2] - Cell area.

% path of the cell centroid

figure(8)

plot(c.stats(:,2), c.stats(:,3));

axis equal

title('Path of the cell centroid')

grid on

xlabel('x coordinate')

ylabel('y coordinate')

% change in cell area

figure(9)

plot(c.stats(:,1), c.stats(:,11));

axis equal

title('Change in cell area');

grid on

xticks(0:5:40);

xticklabels(yticks*c.FI);

xlabel('Time (seconds)')

ylabel('Area (\mum^2)')

Fluorescence stats are slightly more complicated, as there are three channels, but is in the same format as c.stats.

c.fluoStats contains all the data. The dimensions are: (Frame, Channel, measure), where measure one of those listed in c.fluostatsHeaders

% for example

c.fluoStats(:,2,7); % for all frame, channel 2, mean cyto fluorescence

c.fluoStats(1:10,1,11); % for the first 10 frames, channel 1, %age cortex fluo.

The measures computed by ANA are as follows (http://pilip.lnx.warwick.ac.uk/docs/master/QuimP_Guide.html):

Total fluo. [pixels intensity] - Total fluorescence. Sum of all pixel intensities within the cell outline.
Mean fluo. [pixels intensity] - Mean fluorescence. Average intensity of pixels within the cell outline.
Cortex width [microns] - Width of the cortex, as specified by the user.
Cyto. area [microns2] - Area of the cytoplasm (area of the whole cell minus the cortex area).
Total cyto. fluo. [pixels intensity] - Sum of all pixel intensities within the cytoplasm.
Mean cyto. fluo. [pixels intensity] - Average pixel intensity within the cytoplasm.
Cortex area [microns2] - Area of the cortex.
Total cortex fluo. [pixels intensity] - Sum of all pixel intensities within the cortex.
Mean cortex fluo. [pixels intensity] - Average pixel intensity within the cortex.

%plot them against frame number

figure(10);

plot(c.frames, c.fluoStats(:,1,7),c.frames, c.fluoStats(:,1,11));

grid on

legend('Mean cyto fluo','%age Cortex')

title('Cortex fluorescence')

xticks(0:5:40);

xticklabels(yticks*c.FI);

xlabel('Time (seconds)')

ylabel('Pixel intensity')

Tracking through maps

We can track using ECMM data recorded in the co-ordinate and origin maps I have provided all the code to do so

Track forwards.

Typically, the backward tracking gives better results, due to the nature of ECMM mapping.

figure(11);
hold off
plotMap(c.motilityMap, 'm', c.FI); % plot the motility map
trackF = trackForward( c.forwardMap, 1 , 120, 35 );
hold on
plot(trackF(:,2), trackF(:,1), 'r','LineWidth',2);
hold off

Track backwards

trackB = trackBackward( c.backwardMap, 30 , 223, 30 );

hold on

plot(trackB(:,2), trackB(:,1), 'b','LineWidth',2);

hold off

figure(12)

plotMap(c.fluoCh1Map, 'f', c.FI)

trackB2 = trackBackward( c.backwardMap, 40 , 80, 30 );

hold on

plot(trackB2(:,2), trackB2(:,1), 'b','LineWidth',2);

hold off