Screen-set selection

The screen set selection algorithm described here derives from an algorithm developed by Cooper et al. for the creation of sorted dictionaries of words using a form of distribution sorting". ... 

We thank Janey K. Cringean for developing the original version of the screen set selection algorithm described here, the British Library Research and Development Department for funding and the Cambridge Crystallographic Data Centre for its active collaboration in this work. 

In the absence of factual corrosion information for a particular set of fluid conditions, a reasonably good selection would be possible from data based on the resistance of materials to a very simifar environment. These data, however, should be used with some reservations. Good practice calls for applying such data for preliminary screening. Materials selected thereby would reqmre further study in the fluid system under consideration. 

Methods of this nature are adequate for screening sets of hybridomas but not for selection from much larger libraries of antibodies. So, most recently, selection methods employing suicide substrates (Section 7) (Janda etal., 1997) or DNA amplification methodology (Fenniri et al., 1995) have been brought into the repertoire of techniques for the direct identification of antibodies that can turn over their substrate. However, the tedious screening of hybridomas remains the mainstay of abzyme identification. 

Samples that are advanced to the next level of the screen are selected on the same basis as the calibration those inhibiting at least 80% of the tumor cells within two to three standard deviations from the mean for a one-run set. Figure 8.4 shows a typical "scatter plot" for 3680 compounds tested in a run set. The mean inhibition for all the compounds screened was 7.7%. The horizontal bars represent the standard deviations above and below that mean. The 105 compounds deemed "active" (shown by the bold bar in the... 

Fig. 3.8 Typical generic structures selected for the reduced complexity screening set.

A 15-fold glass tube parallel packed-bed reactor was introduced [67, 68], which is similar to conventional catalyst testing equipment (Figure 3.39). The premixed reactant gas is supplied by a 16-port valve either to the bypass or to one of the 15 reactors. By a second 16-port valve, the product gas stream of a selected reactor can be channeled to the quadrupole mass spectrometer. The full automation of the screening set-up allows the investigation of 15 catalysts per day [67]. 

---