Site frequency spectrum analyses for deviations from neutrality: Tajima's D, Fu and Li's F, Fu and Li's D2, etc.
Nucleotide composition, codon bias (if coding).
Choose among selective models using codeml in PAML.
REL/FEL/IFEL/SLAC implemented in HyPhy. FUBAR. Shifts in biochemical properties. The DataMonkey web server will perform these analyses.
Your question is extremely open ended, and it is often the case that you don't want to do analyses "just because." It's more about answering what is important for your system than throwing everything at the wall and seeing what sticks, so to speak. I hope these suggestions will give you some ideas.
I m checking the selection in gene and expecting it is positively evolving. I have applied codon substitution models to estimate ka/ks using different tools. I wanted to know, what kind of selection force is acting on gene in human lineage. Results of ks/ks shows that gene is under negative selection. I suppose, might be there are few positively evolving sites as compare to sites that are under negative selection in gene. Then I applied Tajima's D and Fu and Li test to check weather it deviates from neutrality in population or not. I didn't got significant results for positive selection.
I have review different research papers. they go towards structural side after finding negative selection on a gene.
I am interested in knowing, from evolutionary aspect what are the possible ways for me to the proceed study.
If you are suspecting only few sites in the human gene are under positive selection, a likelihood-ratio test (LRT) using branch-site model in codeml might be an appropriate test and also to find those sites. I found this tutorial very useful for the LRT.