Scheduled PRM for Light-Heavy PQ500 in Plasma with Narrow Acquisition Windows

Analysis of the Heavy Peptides

In step 3 we created two candidate final methods for the 60 and 100 SPD assays. Take the file pq500_60spd_plasma_multireplicate_results_refined.sky and pq500_1000spd_plasma_multireplicate_results_refined.sky, and resave them in the folder Step 4. Plasma Light-Heavy Narrow Window, with names like pq500_60spd_plasma_final_replicates.sky and _pq500_100spd_plasma_final_replicates.sky. In this step we’ll analyze the results of the light/heavy methods created in Step 3. Of particular interest will be the histogram of coefficient of variance values for the peak areas.

Use File / Import Results / Add single-injection replicates in files and press Okay. Select the 10 files in Step 4. Plasma Light-Heavy Narrow Window\Raw\60SPDReplicates and press Open. Remove the common prefix and press Okay to load the results. Remove the PlasmaMultiReplicate with Edit / Manage Results and Save the document.
Select View / Peak Areas / CV Histogram. The CV histograms have ~94% of the targets with CV < 20%, with medians of 3.8 and 4.9%, which are excellent.

You can click on the histogram, which will open a Find Results window with some of the peptides that are close in CV to the value that was pressed. Double clicking any peptide sequence in the Find Results table will make that peptide active, whereupon one can check the peak shape, peak area, and retention time variations for the 8 replicates. Many/most peptides have results like LFGPDLK below.

Another interesting plot is the 2D CV Histogram, found under the View / Peak Areas menu. Here we see that there are 45 targets with CV > 20%. Clicking on the dots makes the peptide in question active. An example peptide SLADELALVDVLEDK is shown below, that has > 20% CV. It has a very skinny peak shape, eluting during the column wash portion of the run. Our analyses show that peptides very early and late in the assays have a high probability of have poor CV. One way to filter these peptides out is based on their narrow peak shape. PRM Conductor can set a minimum LC peak width bounds to do this.

Some peptides have bad CV's because they have interference an interference that varies from one replicate to another, like ANHEEVLAAGK below.

One could simply filter all the peptides with a CV greater than a threshold from the document, using Refine / Advanced / Consistency, and set 20% in the CV cutoff box. Or one could try using MS3 acquisition in PRM Conductor, to save those peptides with poor CV's.

Analysis of Both Light and Heavy Peptides

Now we will add in the light precursors, that were measured but currently are not in the Skyline document. Save the document and then Save again with the names pq500_60spd_plasma_final_lightheavy_replicates.sky and pq500_100spd_plasma_final_lightheavy_replicates.sky.
Use Refine / Advanced, and select the Add box. The Remove label type combo box title changes to Add label type. Select light and press Okay to close the Refine dialog. Each peptide will now have its light precursor added. Use Edit / Manage Results, select all the replicates and press the Reimport button.

Use View / Peak Areas / 2D CV Histogram, and set Normalized to Heavy. The PQ500 standards were designed to look for a set of proteins of interest, not all of which are expressed in normal plasma. For this reason many of the Light / Heavy area ratios are low, and have high coefficient of variance.

Some peptides, like AGALNSNDAFVLK have significant endogenous peptide, and therefore have very good CV. Other peptides like EILVGDVGQTVDDPYATFVK have no observable endogenous peptide, and the area ratio therefore has a much higher CV. This general trend is observed in the CV histogram above.

This is the end of Step 4. We've demonstrated how to analyze replicate data for absolute quantitation with light and heavy peptides. A next step that some users will want to perform is a dilution curve. For absolute quantitation this takes two forms.
- Heavy Dilution: Constant endogenous (light) peptide amount, varying heavy spike-in concentration
- Light Dilution: Constant heavy spike-in concentration, varying endogenous sample.
The Light Dilution is a little easier to perform, because with the Settings / Peptide Settings / Modifications / Internal standard type is set to heavy, and thus Skyline uses the integration boundaries of the heavy peptides to integrate the light signals and determine whether the light/heavy ratios are sufficient for quantitation.
The Heavy Dilution is difficult, because eventually Skyline can't find the heavy peptide signal, and doesn't keep a constant integration boundary. Sometimes Skyline will jump over to the next biggest LC peak and ruin the dilution curve. We have sometimes used a script to set constant integration boundaries and solve this issue.
Calculating LOQs and LODs for large scale assays is still a little difficult, and we have used python scripts to do this. Skyline is also working on making improvements, and there will be updates in the future. We are submitting a paper soon that will have links to these scripts, for the intrepid that might be interested in exploring them.