Motivation:
Two major comparisons are contained in this experiment. The first is a comparison of two DIA methods which differ by window width and resolving power. The second is a comparison of sample prep methods which differ only in whether the samples remain in solution after elution in the SP3 protocol, or whether they are dried down.
Samples:
Pooled human ventricular sample was digested in replicates by Julia (early March 2022). Samples labelled VEP-##A were not dried down at the end of SP3 while samples labelled VEP-##B were speed-vac'd, then resuspended for analysis. 6 VEP-A samples were injected multiple times to compare several DIA methods. Files with 8mz_ were run with 8 m/z wide windows and 15k resolving power. Files with 12mz_ were run with 12 m/z wide windows and 30k resolving power. A 400-1000 m/z gas-phase fractionated library was also collected for peptide detection.
Peptide detection:
All files were searched with EncyclopeDIA (1.12.31). Gas-phase fractionated library was generated from the 4m/z wide windows, 400-1000m/z, searched against a human canonical prosit library ("uniprot_human_25apr2019.z3_nce.dlib") with matching fasta ("uniprot_human_25apr2019.fasta"). Individual injections were searched against the Gas-phase fractionated library ("VEP_GFP_Library_400-1000mz.elib"). Parameters for all searches: normal target decoy, trypsin, CID/HCD fragmentation, 10 ppm precursor, fragment, and library mass tolerance, Percolator v3-01, 8 quantitative ions, min 3. The same sample types were used to generate quantitative reports. Peak area and additional characteristics were extracted in Skyline and exported via document grid. Data processing and searching was performed by Deanna.
Analysis results:
For the first comparison the 12mz window, 30k resolving power method shows improved detections and reproducibility without a decrease in the numbers of points across the peak. For the second comparison there was very little difference in the two sample processing methods (A in-solution vs B dried-down). Slight differences likely due to LC adjustments don't seem to impact the overall detections and reproducibility. Statistical analysis was performed in R by Deanna between 3/21/2022 and 3/29/2022. Results are found attached below.