Abstract Much attention has been focused on RNA transcript bias. The quality of isolated mRNA used for RNA-seq and other downstream analyses can contribute to sample specific bias of transcript abundance. If the library is prepared from polyA enrichment of degraded RNA or with reverse transcriptase using oligo dT priming then 3′ positional bias must be considered (Irizarry 2016). Isolated total RNA is routinely examined for its quality prior to committing to costly gene expression analysis. Since total RNA consists of ~85% rRNA, analysis primarily reflects rRNA quality and the mRNA quality is inferred. Typically, measurements consist of 260/280 nm ratio, 28S/18S rRNA ratio and capillary electrophoresis fragment analysis or gel-based separation and imaging methods. These methods which have become industry standards of quality generate an RNA score such as RNA integrity number (RIN) between 0-10 with >8 being acceptable for RNA-seq. Generating RIN score requires experienced technical ability, specialized equipment, as well as additional cost and time prior to committing to the next step of building the library for sequencing. In this presentation we examine the integrity of the RNA using a combination of nucleic acid specific fluorescent dyes which display different binding specificity depending on the primary and secondary structure of the RNA. Fluorescence signal is analyzed using machine learning based algorithm. For machine learning, a training data set of data was generated, followed by a test set of data used to determine accuracy. Data depicting the actual vs. predicted class are reported in a confusion matrix. Variable of importance analysis was used to determine the factors most important for predicting accurate RNA integrity quality (IQ). With machine learning the initial determination of accuracy using the test data, ranged between 92.7 - 95.7% over the entire range depending on the algorithm. With further refinement of the machine learning approach, RNA IQ had an accuracy of ±0.65 and ± 0.39 standard deviation (SD) for IQ range 0 - 10 using the test data set. The accurate and low SD generated using nucleic acid binding dyes makes for a quick, simple and easy to use assessment of isolated RNA. The dye binding and fluorescent measurement is rapid and requires only a few minutes to generate. Comparison of the dye/RNA based fluorescence based determination of RNA IQ was benchmarked against the industry standards using the BioAnalyzer system (Agilent) to generate a RIN score and other gel analysis methods to determine concordance and utility of this method. Technical ability, cost and time investment are minimal allowing for rapid sample characterization prior to further analysis and commitment to downstream sequencing steps. Citation Format: Scott T. Clarke, Dylan Poulsen, Chris Vonnegut, Debra Gale, Kathy Free. Rapid fluorescence based assessment of RNA integrity quality using machine learning [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2438.