DNA-meter, a calibrated microarray

Step 1: Average and merge

For each experiment, organize microarray signal intensities into a simple text file of two columns, ProbeID  Intensity:

A_67_P05181479   184.6739

A_67_P20444947   300.5581

A_67_P04577152   6376.69

A_53_P127264     37.15217

...              ...

Run average.exe (download | .cpp) to average signal intensities per probe, assign experiment ID and merge experiments into one file and calculate standard errors.

Command

average file1.txt id1 file2.txt id2... > out.txt

id1, id2, etc are experiment IDs; for calibration use concentrations.

Examples

A calibration curve produced from a dilution series experiment (example dataset, output). 

average calib_0_0625.txt 0.0625 calib_1.txt 1 calib_2.txt 2 calib_0_5.txt 0.5 calib_0_25.txt 0.25 calib_0_125.txt 0.125 > calibration_curve.txt

Single experiment (exp1_1_1raw.txt output)

average exp1_1_1raw.txt 1_1_1 > exp1_1_1avg.txt

Step 2: Calibrate microarray

If the microarray is already calibrated, proceed to Step 3.

Run calibrate.exe (download | .cpp), which will use merged averaged signal intensities of the calibration curve and will produce parameters of calibration. Calibration output format is  ProbeID   a   b   R2   CurveID:

A_53_P100398     0.000691116   0.00325183    0.997242  L

A_53_P100736     4.28337   0.296658  0.981139  F

A_53_P102504     0.000593089   0.00524655    0.998836  L

...                                           ...

Command

calibrate  < calibration_curve_file.txt [-T] [-R2 {0.9}] [-freund {F}] [-linear {LN}] [-langmuir {LG}]  > out.txt

Input file produced in step 1.

-T      tabulated output for inspection elsewhere

-R2 {0.9}       R2 cut off, e.g 0.9, provided by user

-freund {F}     Freundlich calibration; output marker, e.g. F, provided by user

-linear {LN}    Linear calibration; output marker, e.g. LN, provided by user

-langmuir {LG}  Langmuir calibration; output marker, e.g. LG, provided by user

When more than one calibration curve suggested, the one with the highest R2 is saved into the output.

Example

Input file is produced by Step 1 (calibration_curve.txt, output)

calibrate < calibration_curve.txt -R2 0.98 -freund F -langmuir L > calibration.txt

Step 3: calculate concentrations

Run concentration.exe (download | .cpp) using averaged signal intensities from Step 1 and calibration parameters from Step 2 to calculate concentrations and standard errors.

Command

concentration  < calibrated_probes_file.txt [-freund {F}] [-linear {LN}] [-langmuir {LG}] -intens sample1.txt sample2.txt...  > out.txt

-freund {F}     Freundlich calibration y=exp(a)*x^b; input marker, e.g. F, provided by user

-linear {LN}    Linear calibration y = a+b*x; input marker, e.g. LN, provided by user

-langmuir {LG}  Langmuir calibration y = Y_max*K*x/(1+K*x); input marker, e.g. LG, provided by user a = 1/(K*Y_max); b = 1/Y_max

Example

Averaged signal intensities for an experiment are produced in Step 1 (exp1_1_1avg.txt) and calibration parameters are produced in Step 2 (calibration.txt).  Intensities are calculated and saved into output.

concentration < calibration.txt –freund F -langmuir L  -intens exp1_1_1avg.txt > conc_1_1_1.txt