ANaMed / BIOAGEModelTrainer · 5 models
Scientific analytics · methods · results

Models. Charts. Scientific evaluation.

What each model estimates, how it is evaluated and what the results show, from predictive performance to calibration and classification performance.

Internal analytical evaluation3 regression models · 2 classifiersResearch use

Scientific status: software checks and primary data

Software corrections have been prepared and tested: fitting with preprocessing learned only on the training data, consistent metric calculation and checks of the original HSE fields. The models have not been retrained on the study data; no new individual predictions or external scientific validation are claimed.

Historical exports and internal evaluations are retained below. A code correction does not retrospectively correct primary responses or establish the reliability of an individual result. Independent platform functions are preserved. Information about prepared code does not confirm that the live API has been updated.

6 720study observations
367prepared features
5research models
3analytical domains
01 /

Results in charts

Different tasks require different evaluation methods. Explained variation is shown for regression; classification performance gives equal weight to each class.

Regression models

R² · reproducing the target measure

R²: internal model evaluationCoreClockScore Beta: 0.990; rTL · FastForest: 0.035; TelomereRiskScore: 0.044. A common scale from zero to one.00.250.50.751CoreClockScore Beta0.990rTL · FastForest0.035TelomereRiskScore0.044
The models have different targets. The high CoreClockScore R² describes reproduction of an age-associated index: age is included among the input features.
Categorical models

Macro-F1 · saved holdout evaluation

Macro-F1: internal model evaluationrTL · three classes: 0.351; TelomereRiskScore: 0.389. A common scale from zero to one.00.250.50.751rTL · three classes0.351TelomereRiskScore0.389
Macro-F1 averages the F1 scores of the three classes with equal weight. Saved holdout results are shown here. The confusion matrices represent separate evaluation data; the differences are explicitly described below.
What these results mean

CoreClockScore describes an age-associated index. Continuous rTL and TelomereRiskScore reflect a small explained component of target variation. For classifiers, overall accuracy, performance within each class and a simple benchmark, the majority-class rule, should be considered together.

02 /

What the five models show

Each model's algorithm, task and output are presented alongside internal evaluation metrics. Values in the cards are rounded; regression targets use different scales, so MAE and RMSE should be compared within a task. Exported metrics and matrix-derived metrics are shown separately for classifiers.

Model 01RegressionLightGBM48 features in the model specification

CoreClockScore Beta

A composite age-associated index for research on age-related profiles.

Internal holdout evaluation
0.990R²
0.019RMSE
0.003MAE

Age is included among the features and has the highest importance estimate. The high R² relates to reproduction of the target index; an independent ageing signal requires separate analysis. RMSE and MAE are expressed on the index scale.

Target measure

CoreClockScore_beta

Input groups

Age, clinical and biochemical measures; imputed values and missingness indicators.

Model 02RegressionFastForest363 features in the model specification

rTL — continuous estimation

A continuous telomere-profile model for research comparisons.

Internal holdout evaluation
0.035R²
8.155RMSE
6.409MAE

R² ≈ 0.035 corresponds to about 3.5% explained variation. MAE ≈ 6.409 and RMSE ≈ 8.155 are points on the analytical rTL_0_100 scale (0–100). The construction of this scale must be distinguished from the laboratory T/S ratio.

Target measure

rTL_0_100

Input groups

Anthropometric, hormonal, clinical-epidemiological and questionnaire measures; missingness indicators.

Model 03ClassificationFastForest363 features in the model specification

rTL — categorical stratification

Three analytical telomere-profile classes: short, intermediate and long.

Saved holdout result
0.351Macro-F1
0.366MacroAccuracy · export
0.9803LogLoss

The export stored MacroAccuracy in the Accuracy field. This is not the proportion of all predictions that are correct. The saved-result metrics and the matrix recalculations below refer to different evaluation artefacts; they have not yet been reconciled to a single run.

Target measure

rTL_class

Input groups

Anthropometric, biochemical, clinical-epidemiological and questionnaire measures, including missingness indicators for hormonal and reproductive features.

Model 04RegressionFastForest39 features in the model specification

TelomereRiskScore — continuous estimation

A continuous telomere-associated index of the age-metabolic profile.

Internal holdout evaluation
0.044R²
0.202RMSE
0.158MAE

R² ≈ 0.044 corresponds to about 4.4% explained variation. Errors are expressed on the TelomereRiskScore scale, a calculated index rather than the probability of a clinical event.

Target measure

TelomereRiskScore

Input groups

Anthropometry, sex, age, education, sleep and behavioural measures.

Model 05ClassificationFastForest39 features in the model specification

TelomereRiskScore — categorical stratification

Three analytical TelomereRiskScore levels: low, intermediate and high.

Saved holdout result
0.389Macro-F1
0.391MacroAccuracy · export
1.1135LogLoss

Target-index levels: below −0.1, from −0.1 to 0.1 and above 0.1. Metrics from the detailed holdout block are shown; recalculation of the saved matrix is considered separately.

Target measure

TRS_Class

Input groups

Waist circumference, sex, age, BMI, sleep measures, education features and missingness indicators.

R² is explained variation relative to the mean; MAE is mean absolute error; RMSE is root mean squared error. Standard Accuracy is the proportion of correctly classified observations; MacroAccuracy is mean class recall (previously mislabelled Accuracy in the export); Macro-F1 is mean class F1; LogLoss evaluates probabilistic predictions, with lower values indicating better performance.

03 /

Calibration profiles

Each point represents the means for one of ten calibration groups. Proximity to the diagonal indicates agreement between the mean prediction and the mean observed value. Model scales differ.

Group meanEquality of prediction and observation

CoreClockScore Beta

CoreClockScore Beta: observed mean and mean predictionTen calibration groups. The dashed diagonal indicates equality of the means. Exact values are listed in the table below the chart.0.210.210.370.370.520.520.670.670.830.83Group 1: prediction 0.212, observation 0.218, n=135Group 2: prediction 0.283, observation 0.286, n=134Group 3: prediction 0.339, observation 0.340, n=134Group 4: prediction 0.393, observation 0.397, n=134Group 5: prediction 0.448, observation 0.449, n=135Group 6: prediction 0.511, observation 0.512, n=134Group 7: prediction 0.594, observation 0.591, n=134Group 8: prediction 0.668, observation 0.666, n=134Group 9: prediction 0.742, observation 0.741, n=134Group 10: prediction 0.829, observation 0.824, n=135Mean predictionObserved mean
10 groups · total n = 1 343. This model's target scale.

Values by group

GroupnPredictionObserved
11350.2120.218
21340.2830.286
31340.3390.340
41340.3930.397
51350.4480.449
61340.5110.512
71340.5940.591
81340.6680.666
91340.7420.741
101350.8290.824

rTL · continuous model

rTL · continuous model: observed mean and mean predictionTen calibration groups. The dashed diagonal indicates equality of the means. Exact values are listed in the table below the chart.28.428.438.538.548.648.658.658.668.768.7Group 1: prediction 28.414, observation 38.211, n=135Group 2: prediction 38.113, observation 44.063, n=134Group 3: prediction 42.501, observation 47.227, n=134Group 4: prediction 46.279, observation 48.786, n=135Group 5: prediction 49.207, observation 49.846, n=134Group 6: prediction 51.856, observation 51.228, n=134Group 7: prediction 54.512, observation 52.976, n=135Group 8: prediction 57.668, observation 54.196, n=134Group 9: prediction 61.435, observation 56.252, n=134Group 10: prediction 68.686, observation 58.141, n=135Mean predictionObserved mean
10 groups · total n = 1 344. This model's target scale.

Values by group

GroupnPredictionObserved
113528.41438.211
213438.11344.063
313442.50147.227
413546.27948.786
513449.20749.846
613451.85651.228
713554.51252.976
813457.66854.196
913461.43556.252
1013568.68658.141

TelomereRiskScore · continuous model

TelomereRiskScore · continuous model: observed mean and mean predictionTen calibration groups. The dashed diagonal indicates equality of the means. Exact values are listed in the table below the chart.-0.50-0.50-0.25-0.250.010.010.270.270.520.52Group 1: prediction -0.503, observation -0.211, n=135Group 2: prediction -0.301, observation -0.149, n=134Group 3: prediction -0.202, observation -0.135, n=134Group 4: prediction -0.123, observation -0.072, n=134Group 5: prediction -0.057, observation -0.043, n=135Group 6: prediction 0.007, observation -0.006, n=134Group 7: prediction 0.082, observation 0.033, n=134Group 8: prediction 0.174, observation 0.096, n=134Group 9: prediction 0.292, observation 0.131, n=134Group 10: prediction 0.524, observation 0.305, n=135Mean predictionObserved mean
10 groups · total n = 1 343. This model's target scale.

Values by group

GroupnPredictionObserved
1135-0.503-0.211
2134-0.301-0.149
3134-0.202-0.135
4134-0.123-0.072
5135-0.057-0.043
61340.007-0.006
71340.0820.033
81340.1740.096
91340.2920.131
101350.5240.305

The charts compare group means and show aggregate agreement between predictions and observations. The distribution of individual errors requires separate analysis.

04 /

What the classifiers show

Two sets of evaluation data

Source-code check: ClassificationTrainer.cs, line 220, writes MacroAccuracy to a field named Accuracy. Thus, 0.366433 is the macro-averaged export metric, not standard accuracy. From the saved rTL matrix, standard Accuracy = 762/1344 = 0.566964, while mean class recall ≈ 0.369696. The latter also differs from 0.366433: the source code uses separate training runs (lines 199, 365 and 134), so the artefacts have not yet been reconciled to a single run. Values are retained unchanged; this is a correction of interpretation, not retraining.

ModelEvidence sourceMetric from that sourceMacro-F1LogLoss
rTLSaved holdout · MacroAccuracy (Accuracy field)0.3664330.3509910.980289
rTLRecalculated matrix · standard Accuracy · n = 1 3440.5669640.353547—
TelomereRiskScoreSaved holdout · MacroAccuracy (Accuracy field)0.3906000.3891001.113500
TelomereRiskScoreRecalculated matrix · standard Accuracy · n = 1 3430.3909160.389128—

All matrices, recall values and majority-rule comparisons below refer specifically to matrix recalculations. Regression fields incorrectly included in the packaged TRS classifier metadata are not used as classification metrics.

Matrices and their recalculated metrics

The matrices show correct predictions and errors by class. Accuracy, Macro-F1, class recall and the majority-class rule are calculated directly from these values.

rTL · categorical model

n = 1 344 · rows: true class; columns: prediction
True ↓
Predicted →
ShortIntermediateLong
Short2422728
Intermediate2469785
Long421441

Recall: short — 8.6%; intermediate — 86.5%; long — 15.8%.
Matrix-derived Macro-F1: 0.354.

TelomereRiskScore · categorical model

n = 1 343 · rows: true class; columns: prediction
True ↓
Predicted →
LowIntermediateHigh
Low18022755
Intermediate56201246
High20232144

Recall: low — 39.0%; intermediate — 40.0%; high — 38.1%.
Matrix-derived Macro-F1: 0.389.

Accuracy and the majority-class rule

ModelAlways predict the most frequent class

rTL · categorical model

Model56,7%
Majority60,0%

TelomereRiskScore · categorical model

Model39,1%
Majority37,5%
A common 0–100% scale. The majority-class rule is calculated from the true classes in each matrix.

Interpreting the comparison

56,7%

rTL: The model's Accuracy is below the 60.0% achieved by the majority-class rule. Recall for the extreme classes is 8.6% and 15.8%; most predictions fall in the intermediate class.

39,1%

TelomereRiskScore: The model's Accuracy exceeds the 37.5% achieved by the majority-class rule by 1.6 percentage points. Recall for the three classes is approximately 38–40%.

This is a descriptive comparison based on matrix data. The statistical significance of the difference and generalisability to other samples require separate evaluation.

How the metrics are calculated

Accuracy = sum of the diagonal / n. Class recall = correct predictions for the class / number of true observations in the class. Class F1 = 2 × correct predictions / (number of true + number of predicted observations in the class). Macro-F1 is the mean of the three F1 values. Majority-rule Accuracy = size of the largest true class / n.

05 /

How internal evaluation was performed

The described workflow includes feature preparation, a held-out evaluation sample and algorithm selection using a prespecified criterion.

  1. STEP 1

    Preparation

    Median imputation, missingness indicators and exclusion lists. A total of 247 feature rules were retained; automatic outlier processing was disabled.

  2. STEP 2

    80/20 split

    Train: 5 374 observations; test: 1 344. Saved seed = 42.

  3. STEP 3

    Algorithm selection

    Minimise MAE for regression; maximise Macro-F1 for classification.

  4. STEP 4

    Result checks

    Holdout metrics, calibration means, confusion matrices and checks of evaluation artefacts.

Data levelCountMeaning
Original cohort6 720The complete original set of observations.
Prepared ModelTrainer dataset6 7185 374 train + 1 344 test.
Source columns261Structure of the original ModelTrainer dataset.
Prepared features367The processed set; subsets differ between models.
Input-schema parameters238The input schema is a separate representation, not a replacement for the number of prepared features.

Evaluation status and interpretation

Internal holdout evaluation describes performance on held-out data from this research dataset. Independent external and clinical evaluation of these five models is not presented.

  • The configuration contains KFold = 5, but the detailed results are labelled Holdout: they should be interpreted as held-out evaluation, not the average of five folds.
  • Exclusion lists and the blacklist are part of feature control; their presence alone does not establish the complete absence of leakage. For CoreClock, a separate analysis excluding chronological age is particularly important. ModelTrainer distinguishes CoreClockScore_beta and CoreClockScore_pct targets; the beta model specification is presented here, without conflating the two configurations.
  • The stated bootstrap uses 1 000 iterations; the composition of resamples and intervals for the main metrics require a separate protocol. No individual confidence intervals are claimed here.
  • Each calibration table and matrix retains its own sample size: n = 1 343 or 1 344. These totals are not replaced with the overall test-set size.
06 /

Why these algorithms were selected

Comparing the saved candidates helps explain the selection criterion. LightGBM has a lower MAE for CoreClock, although FastTree has a slightly higher R². The classifier comparisons below use saved holdout results, not matrix recalculations.

CoreClockScore Beta

Selection criterion: minimum MAE.

AlgorithmR²RMSEMAE
FastForest0.9114250.05732650.0455986
LbfgsPoisson0.8553670.07325450.0460476
LightGBM · selected0.9902750.01899570.00274859
FastTree0.9905990.01867570.00286867
Sdca0.9572680.03981790.0246531

rTL — continuous estimation

Selection criterion: minimum MAE.

AlgorithmR²RMSEMAE
FastForest · selected0.03490468.15496.40946
LbfgsPoisson-1.45330e+561.00072e+292.72968e+27
LightGBM-0.03757558.455586.58218
FastTree-0.05810618.538826.69325
Sdca-4.87536e+0618328.910715.4

rTL — categorical stratification

Selection criterion: maximum Macro-F1.

AlgorithmMacro-F1MacroAccuracy · exportLogLoss
LightGBM0.3317520.3614171.00788
FastForest · selected0.3509910.3664330.980289

TelomereRiskScore — continuous estimation

Selection criterion: minimum MAE.

AlgorithmR²RMSEMAE
FastForest · selected0.04373970.2022980.15845
LightGBM-0.02622050.2095680.163176
FastTree-0.02380550.2093210.163329
Sdca-0.05302150.2122870.16579

TelomereRiskScore — categorical stratification

Selection criterion: maximum Macro-F1.

AlgorithmMacro-F1MacroAccuracy · exportLogLoss
LightGBM0.36920.37061.151
FastForest · selected0.38910.39061.1135

For rTL, the LbfgsPoisson and Sdca candidates produced numerically unstable results; their values are retained, with scientific notation used for extreme values. Errors from models with different target scales must not be compared directly.

07 /

Data quality and feature preparation

Quality control covers completeness, ranges and numeric formats. The table below uses the prepared ModelTrainer dataset: n = 6 718.

MeasureType of checkCountReview
PSQ5iMissing after cleaning6 718Check the variable's availability and role.
HSEMissing after cleaning6 697Check coding and permissible ranges.
HSEOut of range6 622Compare range rules with the original data dictionary.
EstradiolNumeric parsing errors1 534Check laboratory-value formats.
EstradiolMissing after cleaning4 798Clarify the pattern of availability and the imputation rules.
Thigh circumference · ThighOut of range562Check units and source records.

PSQ5i is missing after cleaning in 100% of rows, and HSE in approximately 99.69%. Missingness, out-of-range values and parsing errors may overlap: these counts must not be added to obtain a participant count. Reasons for missing values and their effect on a given model require analysis of the preparation process.

08 /

Molecular measurements: a separate stream

MSP panel: 478 laboratory samples

The laboratory molecular panel and calculated models are presented separately.

rTL_0_100 is an analytical 0–100 scale. Classes and indices are used for research comparisons.

Research models.