Newer
Older
DESCRIPTION
-----------
Model to simulate the spread of Mycobacterium avium subsp. paratuberculosis (Map) within and between herds via trade mouvements between dairy herds.
Control measures can be applied :
- trade mouvements rewiring to foster exchanges between herds of the same status (according to within-herd prevalence) (w option)
- hygiene, by decreasing calf exposure to bacterial environment (e option)
- early culling of high shedders (k option)
Model is mecanistic (infection dynamics) and data-driven (movements and demography), stochastic, discret-time and individual-based.
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
77
78
79
INPUTS
------
Data: BDNI (french database of cattle livestock) extraction from 2005 to 2013 (separator is a tabulation)
- BirthFemaleEvents_withId.csv: number of births per herd per week (time step)
line: 1 per herd
column: 470 -> herd id, then a number per week (52 weeks x 9 years + 1 week)
- CullingRates_withId.csv: death and culling rates per herd per year
line: 1 per herd
column: 82 -> herd id, (death rate at birth, death rate age=0-1 weeks, death rate age=2-8 weeks, death rate age=9-129, culling rate per parity x5) x 9 years
- Headcounts_withId.csv: females number per herd per age group the 1st day of the simulation (2005/01/01)
line: 1 per herd
column: 136 -> herd id, then females number per "week" age (from 1 week old to 130 weeks old), then females number per parity (x5)
- Network_withId_and_breedType_coderace.csv: trade movements between dairy herds
line: 1 per female
column: 6 -> source herd id, destination herd id, time step, age, breed (L for dairy or M for mixte), breed code
Infection initialization in herds is based on simulation outputs from a within-herd model
- selectedInitScenarioForPrev_1to90.csv: animal number per health state per age group according to within-herd prevalence
line: 45001 -> header, then 500 possibilities x 90 prevalences (from 1 to 90)
column: 692 -> animal proportion per health state (S or R, T, L, Im, Ih), then animal proportion per health per age group (5 health states x (130 weeks+5 parities) age groups), then bacteria in 12 different environments (not used)
OUTPUTS
-------
1 file per run per output:
- "twoYearsOldPerHerdPerStep": number of females over 2 years old per herd per time step
- "infected2YOPerHerdPerStep": number of infected females over 2 years old per herd per time step
- "IhPerHerdPerStep": number of Ih per herd per time step
- "positiveIhPerHerdPerStep": number of detected Ih per herd per time step
- "persistencePerHerdPerStep": persistence (0 or 1) per herd per time step
- "statesPerHerdPerStep": herd status (0, 1, 2,...for A, B, C,...) per herd per update (per time step or per year if 'a' option)
- "statusPerHerdPerStep": herd status (0 or 1 for AAA or others) per herd per update (can be used only if 'q' option, which is usually used with 'a' option, so per year)
- "newMoves": 1 line per movement done -> source, destination, time step, age, female health state, source status, destination status, rewiring type
REQUIREMENTS
------------
C++ compiler: g++ (tested version: 7.2.0) => sudo apt install g++
Warning: standard library C++11 is required
tclap (version 1.2.1): sudo apt install libtclap-dev
INSTALL
-------
mkdir results
cd src
make
CHECK (need to install libunittest++-dev)
-------
cd src
make test
make cleantest
LAUNCH
------
cd src
./spreadPTB -h
EXAMPLE for 100 runs, 9 years, no rewiring, low calf exposure and estimated initial conditions:
./spreadPTB -r 100 -e 0.35 -p 0.969
=> time ~7h40 (1 CPU on server, 1 run ~4min30), RAM ~5G, outputs ~10G (8 files per run, ~12M per file)