Introduction to Connectomics: learner worksheet

A 25-minute activity on structural claims, analysis units and study design.

Time: 25 minutes, then 10 minutes of peer review. Work in pairs. Paper and a calculator are sufficient. All numbers below are invented teaching data, not measurements from H01 or another published dataset.

Evidence supplied

A hypothetical EM reconstruction contains four eligible directed neuron pairs. For this exercise, the supplied counts are correct and all relevant contacts within the specified region have been reviewed. Nothing is known about activity or behavior.

A1 to B1: 6 synapses.
A1 to B2: 2 synapses.
A2 to B1: 0 synapses.
A2 to B2: 4 synapses.

The region boundary, cell identities and review status are fixed. A zero means no synapse in this reviewed region, not no connection anywhere in the brain.

1. Name the measurement (5 minutes)

Calculate the total synapse count, fraction of eligible pairs with at least one synapse, mean count per eligible pair, and mean count per connected pair. State each denominator. Why do the last two means differ?

2. Sort the claims (8 minutes)

Use the lecture’s bins: A, supported by the supplied structural evidence; B, requires a stated structure-to-function assumption; C, needs an additional measurement. Give the reason, not just the letter.

A1 has six synapses onto B1 within the reviewed region.
A1 to B1 is physiologically stronger than A1 to B2 because it has more synapses.
A1 causes B1 to fire during a visual task.
A2 never connects to B1 anywhere in the brain.

For claims 2–4, write a defensible replacement and say what evidence the original claim would need. For claim 4, explain why the observation is narrower than the claim.

3. Draft a study brief (12 minutes)

Choose your own question or use: “Do A cells preferentially contact B cells in a defined region?” Supply all of the following:

Question and eligible population, including region boundaries.
Acquisition and reconstruction needed to resolve the proposed analysis unit.
Endpoint, units, denominator and planned sample size. Distinguish a planned sample from the four illustrative pairs supplied here.
Comparison or null model in words. What stays fixed? What may vary?
One error source and a sensitivity check.
A result that would weaken your hypothesis.
An explicit non-claim and the additional measurement needed to address it.

The four-pair example has no alternative target population. It cannot by itself establish preference for B over another type. Do not invent a significance test.

Peer review and exit ticket

Exchange briefs. Underline the endpoint and circle the denominator. Identify one assumption your partner made explicit and one that remains hidden. Revise one sentence.

Submit the calculations, claim classifications and revised brief. On an exit ticket, finish: “My structural result would establish __, but would not establish __.”

Success criteria: calculations have the right denominators; claims stay inside the observed region; the brief includes a comparison, error check and non-claim.

Teaching activity: CC BY-SA 4.0, NeuroTrailblazers. Based on the claim framework in the linked Introduction lecture. Instructor model responses are public; attempt the activity before reading them.