This allows to specificy the number of different OIDs which share a 1k set of positions, per datasets. This implies a dataset with a factor different of one has: * #OIDs = dataset size * factor * #individual positions = dataset size * 1000