Peptide ions elute off the LC column over retention time (x-axis) and separate by mass-to-charge, m/z (depth axis); bar height is MS1 signal intensity. A moving translucent scan window represents the mass spectrometer sampling the eluting peptides. In DDA (Data-Dependent Acquisition) the instrument survey-scans MS1, ranks the peaks it sees, and only sends the Top-N most intense precursors for MS2 fragmentation — low-abundance peptides that never crack the top N in their elution window are silently missed. In DIA (Data-Independent Acquisition) the instrument instead sweeps sequential, fixed-width isolation windows across the whole m/z range every cycle, fragmenting everything inside each window regardless of intensity — near-complete coverage, at the cost of co-fragmented interference when the isolation window is wide.
DDA: select = { p : rank_intensity(p) <= TopN, p in RT_window(t) }
DIA: select = { p : mz(p) in [w_lo, w_lo+IsoWidth], p in RT_window(t) }
Coverage(t) = |identified| / |total peptides|
CV% = CV_base + k1*(IsoWidth-4) [DIA interference]
- k2*(Normalization ON)
+ k3*(Mode == DDA) [stochastic under-sampling noise]
- DDA / DIA buttons — switch acquisition strategy driving which peptides get fragmented each cycle.
- Top-N slider — how many precursors DDA fragments per MS1 survey scan.
- Isolation window — DIA window width in m/z; wider = fewer cycles but more chimeric interference, raising quant CV%.
- Scan / elution speed — how fast the run plays and the scan window advances.
- Normalization toggle — corrects a simulated systematic intensity drift across the run (instrument sensitivity change), lowering apparent CV%.
This mirrors real proteomics pipelines: DDA under-samples low-abundance proteins (high missing-value rate, a major challenge for downstream statistics), while DIA trades some quantitative precision for near-complete, reproducible identification — which is why DIA has become dominant for large-cohort quantitative proteomics.