Percent abundance of each protein was calculated by densitometry of three Coomassie stained gels. Error is standard deviation (n=3). Percent abundance for some proteins was not quantifiable (nq) because they only occurred in mixed spots with other proteins. Asterisks indicate crystallins that were found in both individual spots and mixed spots. Therefore, these percent abundances are underestimates of the true protein abundance. The number of spots containing each crystallin is also indicated. The presence of more than one spot per crystallin is due to truncations or other modifications. Protein accession numbers are for GenBank. Chromosomal location for the corresponding zebrafish and human crystallin genes were identified using zfin and GenBank, respectively. A “?” indicates an unknown chromosomal location; “na” indicates the lack of a human homolog for a specific zebrafish crystallin; and “h” indicates that only one human homolog occurs for multiple paralogous zebrafish genes.