Read this first: PAP-RIDGE organizes PAP records into a practical risk report. It combines the measured PAP, a frozen reference adjustment, previous-PAP trend, and S/D pressure context. Use it to focus review — not to make automatic decisions.
1 What the Tool Is Doing
Clean
Column names are standardized, numbers are parsed, dates are recognized, and questionable entries are flagged.
Bad input can create misleading risk output.
Adjust
Observed PAP is adjusted using a frozen master model for sex, year, curved age, and breed group.
Makes animals more comparable across known sources of variation.
Context
Previous PAP trend and S/D pressure profile are added.
Explains whether the record is stable, rising, unusual, or worth checking.
Score
The adjusted PAP is converted to a latent scale and future PAP probabilities are simulated.
Risk is shown as a gradient, not just high/low.
Review
The report separates decision-facing columns from technical audit details.
Producers can start simple and drill down when needed.
2 What Each Report Tab Means
Test Summary
Overview of the run: number of records, mean PAP, groups present, and model notes.
Start here to confirm the file loaded correctly.
Decision Summary
Main field-facing sheet with ID, PAP, adjusted PAP, risk class, probabilities, trend, S/D profile, and animal summary.
Use this for day-to-day animal review.
High Risk
Filtered list of animals in High or Extreme latent risk.
Use for priority review.
Retest Candidates
Filtered list of Watch or Elevated latent risk animals.
Use to identify animals worth retesting.
Animal Risk Scores
Full animal-level output with every report field.
Use when you need all columns in one place.
Model Details
Adjusted PAP, scaled position, latent PAP, shrinkage, and risk probability internals.
Use for technical review or troubleshooting.
Trajectory and S/D Context
Previous PAP trend, elapsed time, change rate, S/D pressure profile, and notes.
Use when an animal has prior PAP history or unusual S/D values.
Summary by Sex / Breed / Sire
Group counts and average risk summaries by sex, breed, or sire.
Use for family-level review when sire data are present.
Data Quality
Row-level issues such as missing fields, impossible S/D order, or field-note concerns.
Check before acting on surprising results.
3 Core Fields in Decision Summary
mpap
Measured mean PAP from the current test.
The actual observed test result. Do not ignore it.
covariate_adjusted_mpap
PAP after adjusting for sex, year, age, and breed group.
Helps compare animals more fairly.
latent Z / shrunk PAP
Logit-scale transformed score after shrinkage adjustment. Shrunk PAP is the back-converted mmHg equivalent.
The core risk signal. Highlighted in the table — more negative = lower risk, more positive = higher risk.
latent_risk_class
Risk category from simulated future PAP probabilities.
Use as a review category, not a diagnosis.
prob_ge_60 / 70 / 80
Chance simulated future PAP reaches those thresholds.
Higher values mean more concern.
expected_future_pap
Model estimate of future PAP tendency.
Forward-looking context value.
trajectory_class
Summary of previous-PAP history.
Look closely at rising, persistent high, or repeated extreme histories.
pressure_profile_class
S/D context class.
Use to identify unusual pressure patterns or records to check.
data_quality_flags
Possible data-entry or measurement concerns.
Resolve flags before interpreting hard decisions.
4 Key Technical Terms, Plain English
Adjusted PAP
The current PAP shifted up or down based on the frozen reference model. Accounts for sex, year, age, and breed group.
Latent Z / Latent PAP
A transformed PAP score on an unbounded scale. Lets the model work smoothly with probabilities while keeping PAP inside realistic bounds.
Shrinkage
A conservative step that pulls very noisy animal estimates somewhat toward the group average, especially when information is limited.
Shrunk PAP
The latent Z score converted back to mmHg. The value most directly comparable to observed PAP.
Frozen Reference Model
The master adjustment model is fixed and reused for each upload. It is not refit every time a new file is loaded.
Future PAP Simulation
The tool runs many possible future outcomes to estimate probabilities like P(PAP ≥ 70) and P(PAP ≥ 80).
persistence_phi
How much the animal's current PAP state carries forward. Default 0.70 means the current state matters, but some regression toward average is allowed.
Observed PAP Burden
A raw-PAP category based on where the animal falls within the uploaded dataset. Separate from latent risk.
5 Previous PAP and S/D Context
previous_pap
Prior PAP value, submitted or inferred from an earlier row with the same ID.
Shows history and enables trajectory analysis.
pap_change_per_100d
Rate of PAP change scaled to 100 days.
Makes different intervals more comparable.
trajectory_residual
Current PAP minus expected current PAP from history.
Large positive values mean current PAP is higher than expected from history.
pulse_pressure
Systolic minus diastolic.
Wide spread gives hemodynamic context.
sd_mean_pressure_gap
mPAP minus the S/D-derived mean (S + 2D) / 3.
Large gaps can indicate a mean-dominant pressure profile.
Why S and D are not direct risk adjustments: Systolic and diastolic are part of the same pressure test as mPAP. If they were used as direct predictors, the model would mostly be re-predicting PAP from PAP components. The tool uses them as context and quality checks instead.
6 Practical Review Checklist
- Open Decision Summary and sort by latent_risk_class.
- Compare mpap, covariate_adjusted_mpap, and expected_future_pap side by side.
- Check prob_ge_70 and prob_ge_80 as risk gradients — not pass/fail cutoffs.
- If there is a previous PAP, read trajectory_class and pap_change_per_100d.
- Read pressure_profile_class and data_quality_flags before acting on surprising results.
- Use results with animal value, pedigree, environment, production goals, and professional judgment.
📈 Ready to model the genetics?
PAP-SIM projects how your herd mean PAP changes over generations based on bull selection — including latent Z vs. raw PAP strategies.