There are formulae as others have pointed out, but they are really rough estimates that tend to err on the very cautious side (quite rightly). More logical is to work back from an effective concentration, if known. It could be at receptor tissue level in lab tests or animal plasma levels. Then you calculate the probably peak plasma level in a human volunteer assuming 100% absorption (at a rate that is like the animal Pk or a tmax at 30 mins is a reasonable estimate) and match it to that concentration. For first dose to man safety you then take a fraction of that which should depend on the inherent safety of the mode of action of the drug. Some biologics like antibodies are outside this logic although you can use say, 100th rather than a 10th for example because nobody could really criticise you. Then don't make the Tegenero mistake of dosing several volunteers at once.