Flow cytometry is a technique for measuring the protein content of cells. As tens of thousands of cells can be analyzed using this technique in couple of minutes, large size data is produced that is required to be interpreted for clinical and biological applications. Therefore there is essential need to apply computational approaches. I will explain the challenges of the field of flow cytometry data analysis that are interesting from bioinformatics and machine learning point of view including: clustering, features extraction, features selection, and classification. I also elaborate my solutions to address these challenges including: FeaLect (a feature selection method based on the LASSO) and SamSPECTRAL. The later is an enhancement to spectral clustering that had been applied previously to consider the information in eigenvectors of a graph and identify its partitions. Finally, I present our recent results obtained by applying these techniques that were shown to be useful in clinical diagnosis; differentiating between mantle cell lymphoma and small lymphocytic lymphoma.