Fig. 1: Overview of PSO-enhanced thromboSeq. | Nature Protocols

Fig. 1: Overview of PSO-enhanced thromboSeq.

From: RNA sequencing and swarm intelligence–enhanced classification algorithm development for blood-based disease diagnostics using spliced blood platelet RNA

Fig. 1

a, Whole blood is subjected to platelet isolation, total RNA isolation, SMARTer cDNA synthesis, and Illumina TruSeq labeling. Labeled cDNA is subsequently sequenced using Illumina sequencing (e.g., the Hiseq 2500 or 4000 platform), and RNA-sequencing reads can be subjected to the PSO-enhanced thromboSeq pipeline (as indicated by the bird symbol, referring to swarm intelligence). The detailed protocol is described in Steps 1–109. b, Representative examples of platelet isolation. Whole blood is collected in purple-capped EDTA-coated Vacutainers and centrifuged for 20 min at 120g. The platelet-rich plasma (upper yellow layer) separates from the red blood cell layer (lowest layer) and the buffy coat (in between). The platelet-rich plasma is collected in a 15-mL tube and centrifuged for 20 min at 360g. A platelet pellet appears and, following removal of the platelet-depleted plasma, the platelet pellet is resuspended in RNAlater. c, Schematic representation of the PSO meta-algorithm for classification algorithm development. Platelet-RNA-sequencing samples assigned to the training series (top) are used for algorithm development, which is represented as a bin with a bird, referring to swarm intelligence. This results in a dedicated parameter-specific spliced RNA biomarker panel. The performance of this algorithm is assessed in the evaluation series (bottom) and enables PSO to adjust the parameter settings toward those expected to be more stringent and successful (arrow toward the bin). Upon successful parameter selection and algorithm training, the performance is assessed in an independent validation series (lower-right corner). PRP, platelet-rich plasma.

Back to article page