File size: 2,879 Bytes
aecbd27
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
"""Structural check of the three input -> readout chains in MaleCNS v1.0.

Counts direct synapses and a crude two-hop estimate (sum over intermediate
neurons of in-count x out-count, signed by the intermediate's transmitter)
from each stimulated LC population to each descending-neuron readout.
This is a pre-simulation sanity check, not a substitute for the chain tests.

Usage: python -I connectome_paths.py <dir with the three MaleCNS feather files>
"""

import sys

import numpy as np
import pandas as pd
import pyarrow as pa
import pyarrow.compute as pc
import pyarrow.feather as pf

SOURCES = ["LC4", "LC16", "LC9"]
READOUTS = ["DNp01", "MDN", "DNp09"]
SIGN = {"acetylcholine": 1, "gaba": -1, "glutamate": -1, "histamine": -1}


def main(d: str) -> None:
    ann = pf.read_table(
        f"{d}/body-annotations-male-cns-v1.0-minconf-0.5.feather",
        columns=["bodyId", "type", "status"],
    ).to_pandas()
    ann = ann[ann.status == "Traced"]
    nt = (
        pf.read_table(
            f"{d}/body-neurotransmitters-male-cns-v1.0.feather",
            columns=["body", "consensus_nt", "predicted_nt"],
        )
        .to_pandas()
        .set_index("body")
    )
    nt_of = nt.consensus_nt.fillna(nt.predicted_nt)
    traced = set(ann.bodyId)
    type_of = ann.set_index("bodyId").type
    ids = {t: set(ann.bodyId[ann.type == t]) for t in SOURCES + READOUTS}

    w = pf.read_table(
        f"{d}/connectome-weights-male-cns-v1.0-minconf-0.5.feather", memory_map=True
    )
    all_src = pa.array(list(set().union(*(ids[t] for t in SOURCES))))
    all_dst = pa.array(list(set().union(*(ids[t] for t in READOUTS))))
    into = w.filter(pc.is_in(w.column("body_post"), all_dst)).to_pandas()
    out = w.filter(pc.is_in(w.column("body_pre"), all_src)).to_pandas()
    into = into[into.body_pre.isin(traced)]
    out = out[out.body_post.isin(traced)]

    for s in SOURCES:
        for t in READOUTS:
            direct = into[into.body_pre.isin(ids[s]) & into.body_post.isin(ids[t])].weight.sum()
            o = out[out.body_pre.isin(ids[s])].groupby("body_post").weight.sum()
            i = into[into.body_post.isin(ids[t])].groupby("body_pre").weight.sum()
            mids = o.index.intersection(i.index)
            sg = np.array([SIGN.get(str(nt_of.get(m, "")).lower(), 0) for m in mids])
            val = o[mids].values * i[mids].values
            top = (
                pd.Series(val * sg, index=[type_of.get(m, "?") for m in mids])
                .groupby(level=0)
                .sum()
                .sort_values()
            )
            print(
                f"{s:5}->{t:6} direct={direct:6} 2hop_mids={len(mids):4} "
                f"exc={val[sg > 0].sum():.3g} inh={val[sg < 0].sum():.3g} "
                f"top+={list(top.tail(3).index)} top-={list(top.head(2).index)}"
            )


if __name__ == "__main__":
    main(sys.argv[1])