Sparse input representations explain odor discrimination in complex, concentration-varying mixtures

Area of Science:

• Neuroscience
• Olfactory system research
• Animal behavior

Background:

• Animals detect crucial odors amidst complex environmental scent variations.
• Mice can identify target odors in mixtures, but how concentration and background complexity affect this is unknown.

Purpose of the Study:

• To investigate how target odor concentration and background complexity jointly influence odor discrimination in mice.
• To understand the constraints on olfactory perception in naturalistic conditions.

Main Methods:

• Mice trained on a two-alternative forced choice task to identify target odors in mixtures (up to 16 components).
• Systematic variation of target odor concentration after performance stabilization.
• Biophysically grounded model of olfactory bulb glomerular responses and manifold capacity analysis.

Main Results:

• Discrimination accuracy decreased with lower target odor concentration.
• Background complexity had minimal additional impact on discrimination accuracy.
• A linear decoding model matched behavioral performance when neural noise, not background variability, was dominant.
• Neural representations remained structured for odor discrimination across varying background complexities.

Conclusions:

• Olfactory discrimination operates in a noise-limited regime.
• Target detectability is primarily constrained by neural sensitivity.
• Background interference plays a lesser role than neural sensitivity in olfactory discrimination.