- Field of the Invention: The present invention relates to methods and compositions for characterizing the expression patterns of genes and gene families. Specifically, the present invention provides means to generate and monitor gene expression profiles resulting from cellular and physiological changes such that the expression patterns of individual genes or groups of genes can be readily identified and characterized. Background of the invention: Developing methods to detect molecular alterations in biological samples is key to increasing our knowledge about the causes of diseases, the processes of cellular development and differentiation, and other physiological and cellular events, and in developing tools to detect, treat, alter, and monitor these conditions. Perhaps the most significant alteration that can occur in a cell is in its pattern of gene transcription, which exerts profound control on protein levels and activities. The detection of changes in mRNA levels in the thousands of genes expressed by a single cell is an important goal for many research programs. With the extensive amount of cDNA sequence information available through the efforts of genome sequencing projects, as well as those of thousands of individual laboratories, it is becoming increasingly imperative to develop technologies that can utilize this information to study the patterns of gene expression in both development and disease. Most human cancers are the result of genetic changes that result in alterations in the profile of expressed genes within a cell. Methods that can rapidly and accurately measure the expression levels of thousands of genes will play an essential role in furthering our understanding of the causes and nature of progression of human cancers, detecting and monitoring cancers and others diseases, and identifying and developing treatment methods for the diseases. Several approaches have been developed in recent years in an attempt to achieve reliable, economical measurement of patterns and levels of gene expression. These include sequencing-based methods such as expressed sequence tag (EST) databases (See e.g., Adams et al., Nature Genetics 4, 373 ) and SAGE (See e.g., Velculescu et al., Science 270, 484 ), PCR based methods such as differential display (See e.g., Liang et al., Cancer Res. 52, 6966 ; and Liang and Pardee, Science 257, 967 ), and methods based on hybridization to microarrays of EST clones or oligonucleotides (See e.g., Chee et al., Science 274, 610 ; DeRisi et al., Nat. Genet. 14, 457 ; Gress et al., Oncogene 13, 1819 ; Maskos and Southern, Nucleic Acids Res. 21, 4663 ; Pietu et al., Genome Res. 6, 492 ; Schena et al., Science 270, 467 ; and Schena et al., Proc. Natl. Acad. Sci. 93, 10614 ) or by subtractive hybridization (See e.g., Diatchenko et al., Proc. Natl. Acad. Sci. 6025 ).