ABSTRACT

Cells are the fundamental quanta of biology, the basic building blocks of life. There are about 37 trillion cells in the human body (Bianconi et al. 2013) and each cell performs a finite set of functions, depending on its development pathway and environment. The accepted wisdom is that there are between 200 and 500 distinct cell types present in an adult human body (Valentine et al. 1994; Vickaryous and Hall 2006), although a definition

CONTENTS

4.1 Introduction ........................................................................................................................ 123 4.2 Current Approaches to Cell Taxonomy .......................................................................... 125 4.3 Genomic, Transcriptomic, Proteomic, and Metabolomic Tools for Cell

Classification ....................................................................................................................... 127 4.3.1 Single-Cell Genomics ............................................................................................ 127 4.3.2 Single-Cell Transcriptomics ................................................................................. 128 4.3.3 Single-Cell Proteomics .......................................................................................... 130 4.3.4 Single-Cell Metabolomics ..................................................................................... 132 4.3.5 Opportunities and Challenges in Single-Cell ’omics ....................................... 133

4.4 Photonic Tools for Cell Classification ............................................................................. 133 4.4.1 Flow Cytometry ..................................................................................................... 134 4.4.2 Morphological Imaging Techniques ................................................................... 134 4.4.3 Imaging Agent Development ............................................................................... 135 4.4.4 Spectroscopic Imaging Techniques ..................................................................... 136 4.4.5 Live-Cell and Pathology Imaging Systems ........................................................ 137

4.5 Bioinformatics Approaches for Automated Cell Phenotyping ................................... 138 4.5.1 Nonsegmented Approaches ................................................................................. 138 4.5.2 Segmentation Approaches .................................................................................... 139 4.5.3 Data Reduction ....................................................................................................... 139 4.5.4 Unsupervised Learning Approaches .................................................................. 140 4.5.5 Supervised Learning Approaches ....................................................................... 140 4.5.6 Image Analysis Software ...................................................................................... 141 4.5.7 Datasets, Interoperability, and Standards .......................................................... 141

4.6 Challenges and Opportunities ........................................................................................ 142 References ..................................................................................................................................... 144

for what is a cell type, what are the unique characteristics of each cell type, and a systematic survey of all the cells in the human body have not been carried out. Cell types have historically been based on morphological features and qualitative measures of a small set of surface proteins, as well as knowledge of the organ from which the cells were isolated. From these studies, we know that cells in vivo are heterogeneous, even when they are from the same tissue and nominally of the same type, which prompts a key question: Are there a finite number of states in which human cells can exist associated with stable phenotypes? If so, are we now at the stage of being able to map these states and the relationships between them with current technologies in order to provide a more robust taxonomy system for human cells? Can we get to the stage of being able to identify an unknown cell circulating in the human body, for example, locate and characterize a tumor that potentially sheds it?