Neuron Registry Operating Principles

From Program on Ontologies of Neural Structures

Neuron Registry Operating Principles

Draft 1.0 by Giorgio Ascoli, 1/30/10, comments are from David Osumi Sutherland (DOS)

1. Each neuron type in the Neuron Registry is defined by a collection of properties.
   1. The defining properties are all necessary conditions for a neuron to be considered of a given type. This means that if a neuron is known or shown NOT to have one of these necessary properties, then it is of a different type.
   2. Taken together, the collection of properties defining a neuron type are also sufficient conditions for a neuron to be of that type. This means that if a neuron has all of those properties, then it must be of that type.
      • DOS – this is a more ambitious aim than 1.1. If, for any particular class, you (can) only get as far as defining necessary conditions for class membership, this is still useful.
      • GA - Some time stating a set of sufficient conditions may be not more ambitious than do the same for the necessary. For example, one could claim that if you are in CA3, and you are glutamatergic, then you are a CA3 pyramidal cell, just because as of now there are no other known glutamategic cells in CA3. At the same time, having basal/apical dendrites, having spines, and having the soma in the principal cell layer might be considered necessary properties (i.e., if you're aspiny, you not a CA3 pyramidal cell), but not sufficient (there may be basket cells with the same properties).

      Of course, other examples are consistent with your comment. For instance, you could say that being glutamatergic and being is necessary to be considered an entorhinal pyramidal cell, but not sufficient (e.g. you could be spiny stellate cell). So 1.1 and 1.2 appear to be complementary rather than hierarchical. In theory, however, by asking the questions "why are these properties necessary and not sufficient" one should be able to find both the missing properties and the "alternative" neuron types.

      "If a neuron has a subset of the properties defining a particular neuron type, the neuron can be considered of that type if both of the following conditions are met: (i) none of the other properties of the neuron is inconsistent with any of the other properties of that type; and (ii) at least one of its other properties is inconsistent with at least one of the properties of every other neuron type in the registry that also meets all of the same subset of properties (if any)."

   3. It is not required to show that a neuron has all the properties defining a given type in order to conclude that it is of that type. Usually, only a subset of the defining properties is observed in a given lab, time period, and research project. However, stating that a neuron belongs to a type implies the assumption that all of the defining properties would be verified if they were measured.
   4. If a neuron has a subset of the properties defining a particular neuron type, the neuron can be considered of that type if both of the following conditions are met: (i) none of the other properties of the neuron is inconsistent with any of the other properties of that type; and (ii) at least one of its other properties is inconsistent with at least one of the properties of every other neuron type in the registry that also meets all of the same subset of properties (if any).
      • DOS - Not sure I understand the second clause of this. Could you explain further?
      • GA - I realized that it was not clear, but I couldn't come up with better wording so hopefully you can help me out. To keep it simple, suppose that we define a CA3 pyramidal cell as the set of these 5 properties: 1. is in CA3, 2. has spines, 3. is glutamatergic, 4. has basal/apical dendrites, and 5. has soma in principal cell layer.

      Now imagine you find a neuron that is in CA3, has spines, has soma in cell body layer, and expresses the HCN genes. Can we say it's a CA3 pyramidal cell? We first easily verify (i). Then we have to ask the question: what other cells types could also be consistent with these properties? If I find that CA3 chandelier cells are in CA3, have soma in principal cell layer, but are known to lack HCN, they are out. But if CA3 basket cells are in CA3, have soma in principal cell layer, are not *necessarily* aspiny, and it is not known whether they express HCN, then you are forced to say that your candidate neuron is either a CA3 pyramidal cell or a CA3 basket cell.

      "If among the defining properties of the neuron type are the locations of the soma and/or dendrites, the name of that type should begin with the noun describing the region of the nervous system encompassing those locations (e.g. spinal cord motoneuron, dentate gyrus granule cell)."

   5. The Neuron Registry cannot list all of the known properties of a given type, nor is it expected that every neuron type entry in the Registry will have an exhaustive or even extensive list of properties. However, each neuron type entry will list and denote all of its defining properties, and may be enriched with any number of additional properties.
      • This is good. It is basically a statement of the often-used ontology design principle known as ‘minimal commitment’.
2. Each neuron type in the Neuron Registry is associated with a unique identifier (i.e. a code or string of characters not associated with a different type).
   1. The unique identifier is not to be confused with the common name of the neuron. However, each neuron type will also be given a preferred name and possibly one or more synonyms.
   2. If among the defining properties of the neuron type are the locations of the soma and/or dendrites, the name of that type should begin with the noun describing the region of the nervous system encompassing those locations (e.g. spinal cord motoneuron, dentate gyrus granule cell).
      • DOS - A reasonable aim, but you might want to be careful about boxing yourself in with naming rules – especially where they prevent you from using a commonly understood name instead.
      • GA - I agree, this is just meant as a recommendation and not a strict rule, but we've found it useful within the NIF cell working group so far. Also generous use of synonym can be helpful.
   3. If the name of the neuron includes additional terms to discriminate different subtypes, those should be appended at the end (e.g. spinal cord motoneuron alpha, spinal cord motoneuron gamma).
3. The properties defining a neuron type should be expressed, if possible, with a relation and a value, specifying whenever applicable the part of the neuron they refer to.
   1. For example one of the properties defining dentate gyrus granule cells is that their dendrites are in the molecular layer. This is expressed in the Neuron Registry with the relation “located_in”, the value “molecular layer”, and the assignment to “dendrites”.
   2. Every assignment of a defining property to a neuron type must be accompanied by at least one citation of a peer-reviewed publication.
   3. The assignment of a property to a neuron type may be accompanied by notes, e.g. of explanations or comments.
      • I strongly endorse these last 2 points. Many problems with clarity and maintainability of ontologies would be addressed if everyone did this.