Updated publication reference for DOI(s): 10.1021/acs.jproteome.1c00796. Data-independent mass spectrometry is the method of choice for deep, consistent and accurate single-shot profiling in bottom-up proteomics. While classic workflows required auxiliary DDA-MS analysis of subject samples to derive prior knowledge spectral libraries for targeted quantification from DIA-MS maps, library-free approaches based on in silico predicted libraries promise deep DIA-MS profiling with reduced experimental effort and cost. Coverage and sensitivity in such analyses, however, is limited, in part, by large library size and persistent deviations from experimental data. We present MSLibrarian, a workflow and tool to obtain optimized predicted spectral libraries by the integrated usage of spectrum-centric DIA data interpretation via the DIA-Umpire approach to inform and calibrate the in silico predicted library approach. Predicted-vs-observed comparisons enable optimization of intensity prediction parameters, calibration of retention time prediction for deviating chromatographic setups and optimization of library scope and sample representativeness. Benchmarking via a dedicated ground-truth-embedded species mixture experiment and quantitative ratio-validation confirms gains of up to 9 % on precursor and 7 % protein level at equivalent FDR control and validation criteria. MSLibrarian has been implemented as open-source R software package and, with step-by-step usage instructions, is availabe at https://github.com/MarcIsak/MSLibrarian.