Molecular spectroscopy for atomic spectroscopers My experience is with atomic spectroscopy of alkali atoms, I've recently been asked by a friend to help with some advice on analyzing molecular spectrophotometry data in the context of molecular biology, 
To fit spectroscopic data to extract density in atomic spectroscopy you typically need to know the relative transition strengths and natural linewidths for all transitions (for hyperfine resolution spectroscopy), know the interaction path length, know the doppler broadening coefficient (and know the temperature), construct a voigt profile and fit. 
If I wanted to fit spectroscopic data to extract, say, ATP concentration from a spectrophotometer sweep, would I be able to do this by determining the corresponding molecular parameters? Are there caveats to single transition molecular spectroscopy that I should be aware of? It would seem that doppler broadening would be the largest contribution to the linewidth so it I may be able to fit to a simple gaussian profile instead of the more computational intense voigt.
In atomic spectroscopy you usually measure a change in coherent beam intensity so the contribution from fluorescence of the atom can be made negligible in the low power unsaturated regime with polarization analyzers and other small trade-tricks. I'm not completely certain how spectrophotometers are set up with that regards, would fluorescence near the measurement frequency be something I need to take into account?
 A: High-resolution atomic spectroscopy and spectrophotometry of biomolecules in solutions are two different worlds. Large molecules usually don't have well-defined symmetry, so their spectra are broad and structure-less. They are additionally broadened by interaction with solvent molecules (this will be by far the largest contribution to line broadening and shifting).
Instead of transition strengths, people use extinction coefficients $\varepsilon$, the measure of absorption rate by a molecule at a specific wavelength (in specific solvent, at specific temperature etc). If you know this coefficient, you can measure how much light is absorbed and calculate the concentration of a molecule using Beer-Lambert law $$-ln\left ( \frac{I}{I_0} \right ) = \varepsilon lc $$
Since there are so many factors that interfere with such measurements - like other absorbing species, one have to invent some tricks. For example, there is the method of Standard addition, where you add known amounts of measured molecule to your sample and measure the increase in absorption, then extrapolate the plot to find the initial concentration.
