Thanks for the detailed explanations.

Adjusted p-value otherwise called q-value serves to streamline the results of p-values, especially in multiple testing or comparison and the control by the false discovery rate (FDR). That is to say, if p-values suggest about 200 items to have passed a test, q-value will screen the 200 and bring out the real false positives among them. Whatever the case and whenever necessary try to interpret the two values side by side to make your analysis more robust and conclusions stronger and more reliable. For details, please go through the suggested link. http://www.nonlinear.com/support/progenesis/comet/faq/v2.0/pq-values.aspx#:~:text=Another%20way%20to%20look%20at,result%20in%20fewer%20false%20positives.

Either simple provide the complete list or -if you want to pick candidates for your further studies- sort the list by LFC or by p-value and go through it from top to bottom until you find some biologically interesting candidate.

If you wont a subset of the list to be presented in some printed table, sort the list by LFC of p-value and then give as many from the top of the sorted list as you have space.

Cutting a list by some cut-off is never a good idea, particularily not if this part should be used to declare "differentially expressed" genes or just in subsequent analyses (like over-representation analyses etc.).

Using p-values to test hypotheses about differential expression makes no sense, because the study is hypothesis-generating. Look at the top of the sorted list and go down until you get an idea about some testable hypothesis. The perform an approriate experiment, get the data, and test that hypothesis.

A final note: surely, some people will tell you to calculate the false-discovery rate (FDR; look up Wikipedia if required). This requires independent tests - but the differential expression of the genes in a cell is by definition NOT independent! The same applies for other usual methods for multiplicity correction.

Thanks for the detailed explanations.