How do I optimize a direct ELISA?

There are several factors one should consider when optimizing a direct ELISA. 

1. One factor to consider is to to prepare different concentrations of the capture antibody in the buffer, as it is important to ensure too much or too little of capture antibody isn't being added to avoid inaccurate results. 
2. The signal-to-noise ratio should then be checked to identify the concentration which provides the highest ratio. 
3. Another factor is making different blocking solutions for the blocking buffer. If a blocking solution is not preformulated, one should use different concentrations of the protein. Again, the signal-to-noise ratio should be checked. 
4. An additional factor is preparing different concentrations of the detection antibody, followed by checking for strong signal-to-noise ratio. 
5. For optimizing the sample concentration, one should try to use a standard diluent as similar as possible to the matrix of the membrane. If the matrix is not identical to the diluent, different standard diluent solutions should be tested. 
6. The dynamic range for the standard curve should then be checked to see if it is sufficient. If the dynamic range is insufficient, a different diluent should be used. 
7. The linearity of the dilution should also be checked. If a sample has poor linearity, linearity-of-dilution or spike-and-recovery procedures should be carried out. 
8. For optimizing the standard dilution, different concentrations of the enzyme conjugate should be prepared so that it is compatible with the substrate. The signal-to-noise ratio should also be checked. 
9. For optimizing signal detection, a substrate should be used based on its ability of the antigen to be detected with the right tools. If the antigen is able to be detected over a wide dynamic range, the substrate is sufficient. If it is not able to be detected, a more sensitive substrate should be used. 
10. To optimize the enzyme conjugate, different concentrations should be prepared and checked for strong signal-to-noise ratio.