Evaluation of density gradient separation methods.

A new method for evaluating the effectiveness of isopycnic separation processes is developed and explored. This technique is designed to measure the ability of a separation technique to attain the goals of a sorting process. The performance of a separation process is expressed in terms of an inconsistency number, I, whose value ranges from zero to unity. Processes which function perfectly are characterized by an I value of zero. Separation processes which merely subdivide a sample are characterized by an I value of unity. The validity of using I to measure the performance of separation processes was established by demonstrating that alterations in the performance of a well characterized separation technique were fully reflected by changes in the I value for the process. The capacity of four different techniques for ordering red cells with respect to their density was then shown to lie in a definite hierarchy. This hierarchy can be expressed as: Murphy technique (I = 0.21) greater than albumin density gradient (I = 0.43) greater Stractan density gradient (I = 0.53) greater than Percoll density gradient (I = 0.73). The performance was found to be dependent on the exact operating conditions employed and the specific fraction of red cells isolated.