Back

The cost of multiplexing: PFC integrates multiple sources of information in non-orthogonal components accounting for behavioral variability

Roig, J. L. A.; Di Bello, F.; Hassen, S. B. H.; Astrand, E.; Hamed, S. B.

2022-12-29 neuroscience
10.1101/2022.12.28.522139 bioRxiv
Show abstract

The frontal eye field (FEF) is a cortical area classically associated with spatial attention, perception, and oculomotor functions. FEF exhibits complex response properties through mixed selectivity neurons, allowing a high dimensional representation of the information. However, recent studies have shown that FEF encodes information in a low-dimensional regime hence limiting the coding capacity of the neural population. How the FEF encodes multiple sources of information with such limited encoding capacity remains elusive. To address this question, we trained two macaques to perform a visual attention task while we recorded FEF neuronal activity using multi-contact electrodes. FEF neurons encoded task- (time in the trial; CTOA) and behaviour- (reaction time, RT; focus of attention, TA) related parameters prior to the target onset. We found a clear modulation of the RT and TA as a function of the CTOA. Using dPCA, we characterized the functional relationship between neural populations associated with each parameter and investigated how this functional relationship predicts behaviour. We found that CTOA variability was associated with two different components the activation of which was correlated with the TA and the RT, respectively. These CTOA-related components were non-orthogonal with the RT and TA-related components, respectively. These results suggest that, when different sources of information are implemented during task performance, they show a very precise geometrical configuration in non-orthogonal components, which allows a high capacity of information coding at a cost of modulating both the capacity of the monkey to use attention information and its responsiveness toward external stimuli.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

1
The Journal of Neuroscience
1025 papers in training set
Top 1%
12.3%
2
eLife
5828 papers in training set
Top 9%
10.5%
3
Proceedings of the National Academy of Sciences
2444 papers in training set
Top 4%
9.3%
4
Nature Communications
5641 papers in training set
Top 22%
7.5%
5
Journal of Neurophysiology
302 papers in training set
Top 0.6%
6.9%
6
Progress in Neurobiology
47 papers in training set
Top 0.1%
6.9%
50% of probability mass above
7
Scientific Reports
3612 papers in training set
Top 14%
6.0%
8
Communications Biology
993 papers in training set
Top 1%
6.0%
9
Cell Reports
1498 papers in training set
Top 7%
5.2%
10
iScience
1154 papers in training set
Top 5%
3.8%
11
PLOS Computational Biology
1863 papers in training set
Top 9%
3.8%
12
Science Advances
1243 papers in training set
Top 12%
3.1%
13
Cerebral Cortex
396 papers in training set
Top 3%
2.3%
14
PLOS Biology
486 papers in training set
Top 3%
2.3%
15
eneuro
439 papers in training set
Top 4%
2.3%
16
European Journal of Neuroscience
189 papers in training set
Top 2%
2.0%
17
Journal of Cognitive Neuroscience
135 papers in training set
Top 1%
1.3%
18
Frontiers in Systems Neuroscience
22 papers in training set
Top 0.2%
1.1%
19
NeuroImage
903 papers in training set
Top 5%
1.1%
20
Current Biology
665 papers in training set
Top 9%
1.0%
21
The Journal of Physiology
150 papers in training set
Top 2%
1.0%
22
Nature Neuroscience
252 papers in training set
Top 5%
0.8%