What are some ways of differentiating bot transactions from normal users?

I'm looking into a token launch that caused some very abnormal activity on Base network (spiked tx fees over 500x shortly after launch, caused some connection issues). There was speculation that the team used bots to manipulate the volume of their token, and perhaps got into a gas war with MEV bots, or some sort of price oracle manipulation.

I'm not entirely sure of how to go about this part of the investigation.

Considering heuristics like finding average gas paid/tx frequency/number of failed txs, then comparing these factors to a token launch that I know was organic. I'm well versed in onchain data but don't have much experience with bots or ways a team might utilize them or try to obfuscate their use.

Any insight on factors to look for in determining the extent that transactions might have been automated or methods/tools that could make this easier would be appreciated.