Heiko's projects brainstorming

From Vision Wiki
Jump to navigation Jump to search

Pietro's and David's brainstorming - Oct 18, 2006

Possible new projects:

Prediction of aggression

  • Given the 3 seconds leading to an lunge, could one predict the lunge that is coming? What is the spatial configuration of 2 flies that `triggers' the lunge? Can we reduce this to vision, proximity, other sensory information?
  • Does the frequency of this configuration change with aggressive and non-aggressive flies?
  • Or is the configuration leading to aggression different?

Analysis of genetic and environmental influences on aggression

  • Given that a higher proportion of socially isolated flies show lunges than do group-housed flies, is this accompanied by any other changes in behavior? Increased locomotor activity? Generally increased social interactions (e.g., increased male-male mating)?
  • Given Liming's mutations that increase aggressiveness, do these mutations have a similar influence on behavior as does social isolation?
  • Recent data published in Science show that social isolation vs. group housing also influences the frequency of "sleep" (defined as a 5 minute period without activity) during the daytime. Given this observation, is there a link between less sleep and more aggression? Can we monitor sleep during the same periods that we are monitoring aggression?

General analysis of behavioral transitions

  • Given an episode of analysis with multiple behaviors, can we automate analysis of the probabilities of transitions between different behaviors, to construct an ethogram?
  • If so, we can use this method to ask whether manipulations of genes, neurons or environment changes
  1. the probability of transition between two behaviors;
  2. the "vector map" of preferred behavioral transitions

Discovering new behaviours

  • Can the program "discover" new behaviors that would not be apparent to human observation, or patterns of behaviors that would not be apparent?

Correlations between behaviours

  • Sex and fighting: do manipulations that increase courtship frequency also increase aggression, and vice-versa? Do all manipulations that increase male-male aggression also increase male-male courtship?

Cooperativity

In humans there are `group' behaviors: if the likelihood that someone sits on a long low wall (or bench) is much higher if someone else is sitting on it already. Can we see the same behavior in flies? E.g.:

  • Feeding: is it more likely that a new fly will come and eat if another fly is eating already? (Controls needed: control for population hunger, change the total n. of flies in group etc)
  • Fleeing: if one fly takes flight, are more flies likely to take flight as well -- e.g. when a predator looms.