CSF-Seq enables transcriptome-wide profiling of cerebrospinal fluid and identifies a prognostic signature of leptomeningeal disease

This collection of data accompanies the study: Umeh Garcia et al. CSF-Seq enables transcriptome-wide profiling of cerebrospinal fluid and identifies a prognostic signature of leptomeningeal disease. (2026). Data generated as part of this study include transcriptome-wide RNA sequencing of cell-free RNA (cfRNA) isolated from human cerebrospinal fluid (CSF) using CSF-Seq, an optimized workflow for whole-transcriptome sequencing of low-input, fragmented CSF cfRNA. The study includes 129 CSF samples collected from patients with diverse neurological and oncologic conditions, including: Leptomeningeal disease (LMD) from breast cancer and lung cancer Glioblastoma (GBM) Traumatic brain injury (TBI) Non-cancer neurological Brain Tissue RNA (positive control) Sequencing libraries were generated using optimized RNA extraction, library preparation, and deep sequencing to enable transcriptome-wide profiling of coding and non-coding RNA from clinically obtained CSF specimens. Uploaded data include: Processed sequencing data Gene-by-sample raw count matrix generated following alignment and gene quantification Sample metadata Sample annotation file containing de-identified biospecimen and experimental metadata, including disease group, collection modality, patient identifier, and relevant sequencing information Raw sequencing data exceed the storage limit and are pending upload to GEO. FASTQ files for all 129 CSF samples These data support all transcriptomic analyses presented in the accompanying manuscript, including quality assessment, differential expression analysis, pathway enrichment analyses, disease-specific transcriptional profiling, development of the collection method–independent LMD gene signature, and prognostic analyses.