Percussion is typically the hardest group of instruments to EQ. Not only is there an array of sounds created by instruments of all tones and timbres, but they are usually recorded from a slight distance, so a lot of the sounds bleed together when multi-tracking. With cymbals, for instance, usually recorded with overhead mics, it's best to cut the frequencies to below 10K or so, which will help remove most of the nearby drum sounds. When EQing the bass drum, it's a good idea to boost the 100Hz area while cutting the muddier 250Hz area, which helps remove the tom sounds from above. The snares and the toms have more of a punchy tone and so, for them, the engineer can accent those mid-range frequencies avoided by the bass, drum and cymbals.
There are countless ways to EQ the bass, all of which depend on the particular sound that the musician or engineer is seeking. For the sake of a general guide, there are a few tips to follow. First, it's usually best to cut frequencies instead of adding them. The bass is already so powerful that boosting it can result in overkill. Also, unlike the bass drum, the 100Hz area is too low a frequency, making it difficult to distinguish any notes. So it makes sense to cut everything below it. There is also usually a varying unpleasant mid-range frequency that occurs when recording bass. This should be pinpointed and cut to ensure a clean-sounding track.
The guitar may be one of the most dynamic instruments, especially in modern day rock music, but one that should require the least amount of EQing. If anything, removing some of the ultra-high frequencies is advisable. This 15-20k area only tends to emphasize the sound of fingers sliding against the strings between chord changes, which on acoustic guitars especially can be off-putting. The bass EQ also needs to be taken into consideration. For instance, if the bass contains a high amount of mid-range, it's a good idea to remove some of the same mid-range frequencies in order to avoid the two instruments stepping all over each other.
Similar to bass, and even more so, the EQ for vocals can go a number of ways, depending on the person singing as well as the type of song recorded. On most recordings, one basic tip is to remove all the extra low frequencies, from 150Hz and below. Any sound in this range is most likely a mistake, such as an off-timed breath or a harsh "P." The same goes for the ultra-high frequencies. This 3K and up range usually only emphasizes the "S" sounds in vocal takes, as well as voice cracks and heavy breathing. While a good microphone will prevent any need for boosting frequencies, a flat recording may require the engineer to boost the 1Hz area where most vocals hover.