Rank-In

Rank-In harmonizes microarray and RNA-seq transcriptomic data to correct nonbiological variability and enable integrated differential expression and classification analyses.

Key Features:

• Correction of Nonbiological Effects: Mitigates discrepancies and nonbiological variability between microarray and RNA-seq platforms to enable combined analysis.
• Validation and Performance: Validated on public cell and tissue samples including two SEQC reference samples where it perfectly classified 44 profiles and achieved 0.9 accuracy in predicting TaqMan-validated differentially expressed genes (DEGs).
• Discrimination Capability: Distinguished every Glioblastoma (GBM) and heterogeneous colon cancer profile from normal controls in benchmark studies.
• Robustness Across Profile Sizes: For mixed sequence-array GBM profiles of varying sizes, reproduced a median DEG overlap of 0.74–0.83 among top genes, outperforming other methods that did not exceed 0.72.
• Support for Diverse Sample Types: Enables integration of large or small, paired or unpaired, and balanced or imbalanced sample collections.

Scientific Applications:

• Mixed-platform integrative analysis: Enables combined analysis of microarray and RNA-seq datasets for differential expression and classification tasks.
• Cancer transcriptomics and clinical cohorts: Supports integrative studies in cancer (e.g., GBM, colon cancer) and can reduce the sampling space required for clinical patient studies.
• Cross-platform validation: Facilitates validation of DEGs across platforms including comparisons to TaqMan assays and SEQC reference samples.

Methodology:

Rank-In applies an algorithmic harmonization to correct nonbiological variability between microarray and RNA-seq platforms.