I'm sure many people know there are a large number of prediction softwares' available online, with most including high throughput and low throughput validation. Most of these have a filter list, which might make you rethink the best miRNA target results, i.e. TarBase v8? Does anyone have good recommendations or guidelines on which one to prioritise?

1. Prioritising the highest predictive value (i.e. 0.0-1.0)

2. Prioritising the highest high throughput experimental value (0+)

3. Prioritising the highest low throughput experiment value (0+)

4. Combining the highest predictive value that has a high experiment throughput value also

5. Combining the highest predictive value that has a low throughput experiment value also

6. Combining the highest predictive experiment value that has both a high throughput value and low throughput experiment value

7. Combining the highest high throughput experiment value that has a low throughput experiment value and not including the predictive value

If I understand the topic correctly, high predictive algorithms score don't necessarily mean a biological interaction, high throughput experimental values in a kidney don't necessarily mean that the miRNA interaction that I'm looking at will be the same, etc.?