Additional file 1 of Plasma lipid profiling for the prognosis of 90-day mortality, in-hospital mortality, ICU admission, and severity in bacterial community-acquired pneumonia (CAP)
Bibliographic record
Abstract
Additional file 1: Figure S1. Coefficient plot shows a relative concentration of 20 metabolites involve in the discrimination of non-survivors from survivors by DI-MS/MS. Table S1. Unpaired t-tests show 32 metabolites with significant changes [(FDR < 0.05), highlighted in blue] between in-hospital deaths and survivors detected in plasma using DI-MS/MS. Figure S2. Permutation test (200 times) of OPLS-DA model of metabolites in plasma obtained by DI-MS/MS, to validate the predictability of the model to separate survivors (n = 75) from non-survivors (n = 75). The test shows the prediction is valid. Figure S3. PLS-regression shows a very strong relationship between the most differentiating metabolites (n = 20) in the separation of survivors and non-survivors in plasma detected by DI-MS/MS. Figure S4. Coefficient plot shows the relative concentration of increased and decreased metabolites in the in-hospital mortality versus survivors (> 90 days) based on the DI-MS/MS data. Table S2. Unpaired t-test shows 65 metabolites significantly changed [(FDR < 0.05, highlighted in blue) between in-hospital deaths and survivors (> 90 days) detected by DI-MS/MS of plasma. Figure S6, PLS-regression analysis shows a very strong correlation between most differentiating metabolites separating survivors from in-hospital deaths (using the most differentiating metabolites, n = 22) detected by DI-MS/MS of plasma. Figure S7. Permutation test (200 times) to validate the predictability of the model to separate in hospital deaths (n = 26) from survivors (n = 75). The test shows the OPLS-DA prediction is valid. Figure S8. Partial least square regression (PLSR) analysis shows that the most differentiating metabolites obtained by DI-MS/MS are in strong relationship with APACHE III and PSI scores, showing metabolites are highly correlated with paraclinical features. A: APACHE III, B: PSI at day 0, C: PSI at day 1, and: PSI no Age. Table S3. Logistic regression of APACHE III and PSI to predict mortality. This table shows that severity scoring systems (APACHE III and PSI) are not as good as metabolites biomarkers for the prediction of mortality in these cohorts because they have lower sensitivity, specificity, AUROC, p values and regression compared to metabolomics lipid profiling. Table S4. Characteristics of bacterial CAP ICU patients vs. ICU ventilated controls. ŧ APACHE III, ŧ ŧ APACHE II. Figure S9. OPLS-DA of NMR metabolites show a very predictable model to separate the bacterial CAP ICU patients (n = 41) from ICU ventilated control (n = 31) using 1H-NMR, R2Y = 0.871, Q2Y = 0.777, and p = 1.97× 10− 21. Figure S10. OPLS-DA of GC-MS features show an excellent prediction model to separate the bacterial CAP ICU patients (n = 41) from ICU ventilated control (n = 31) using GC-MS, R2Y = 0.915, Q2Y = 0.852, and p = 1.3× 10− 16. Table S5. Summarized OPLS-DA models of two platform to separate ICU-admitted CAP pneumonia (n = 41) from ICU ventilated controls (n = 31), Both NMR and GC-MS show an excellent prediction models as verified by Q2Y, p value, sensitivity, specificity and AUROC parameters. (AUROC = area under receiver operating curve). Table S6. Unpaired t-test shows 22 known metabolites obtained by DI-MS/MS significantly changed (p < 0.05, while 13 metabolites highlighted had an FDR < 0.05) between ICU and non-ICU CAP patients. Table S7. Unpaired t-test shows 37 known features significantly changed (p < 0.05, while 12 metabolites highlighted had a FDR < 0.05) between patients with severe pneumonia (PSI 4 and 5) vs. patients with less severe pneumonia (PSI 1, 2 and 3). Figure S11. Heatmap analysis shows a separation between patients with more severe (PSI 4 and 5) and less severe (PSI 1, 2 and 3) CAP using the most differentiating metabolites by DI-MS/MS. Table S8. Unpaired t-test shows 27 metabolites obtained by NMR significantly changed (highlighted FDR < 0.05) between ICU-admitted bacterial CAP and ICU ventilated controls. Table S9. Unpaired t-test shows 46 known features obtained by GC-MS significantly changed (FDR < 0.05) between ICU-admitted bacterial CAP and ICU ventilated controls. Table S10. Comparison of three analytical platforms shows that DI-MS/MS is more predictive and significant for 90 day mortality than GC-MS and NMR. Table S11. The characteristics of CAP patients who were admitted to ICU vs. CAP patients who were not admitted to ICU.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.029 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.827 | 0.120 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".