Repeated GC-IMS Measurements of Pooled Urine for Orthogonal Projection-Based Signal Correction

This dataset contains repeated gas chromatography–ion mobility spectrometry (GC-IMS) measurements acquired from a pooled human urine sample. The data were generated to investigate signal instability arising from acquisition-related technical variability and to evaluate correction strategies based on orthogonal projections. A total of 135 GC-IMS measurements were collected across nine measurement sessions (batches). In each session, 15 aliquots of the same pooled urine sample were analysed under identical experimental conditions. Because all measurements originate from the same biological source, no biological variability is expected; therefore, systematic differences between measurements can be attributed to technical factors related to the acquisition process. Two acquisition-related variables were associated with each measurement: the batch index (corresponding to the measurement session) and the elapsed time before injection (estimated from the position of the sample in the acquisition sequence). These variables capture between-batch and within-batch technical variability, respectively. Urine samples were originally obtained from healthy volunteers and combined to form a pooled sample used for repeated measurements. The study protocol was approved by the Ethics Committee of Hospital de Reus (approval no. 074/2018). Measurements were performed using a FlavourSpec® GC-IMS instrument (G.A.S., Dortmund, Germany). Each acquisition produced a two-dimensional signal matrix where the axes correspond to gas-chromatographic retention time and ion mobility drift time, and signal intensity represents the detected ion current. Raw instrument files are provided in the original .mea format. These data are intended for methodological research in GC-IMS preprocessing and signal correction, particularly for studying batch effects, acquisition-order effects, and the evaluation of orthogonal projection-based correction approaches. The dataset may also be useful for benchmarking preprocessing pipelines, evaluating feature stability, and testing chemometric or machine-learning workflows applied to GC-IMS metabolomics data. Additional information regarding file structure, acquisition metadata, and annotation variables is provided in the accompanying README file.