Why Science? | Singapore Phenome Centre

Phenotype reflects genes, development, environment and behaviour. Metabolic measurements can capture current state, but diet, medicines, activity, microbes and time make that state dynamic.

The metabolome is dynamic. Food, medicines, activity, microbes, time of day and sample handling can shift measured molecules, so study design and pre-analytical records are essential. The practical capability is phenotype literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Phenotype links biology with lived conditions” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A molecular snapshot is not a complete description of health.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Next chapter

Metabolites include substrates, intermediates and products of cellular and microbial chemistry. Chemical diversity and concentration range mean no single method measures everything.

Mass spectrometry and NMR observe chemistry through different physical signals. Their strengths are complementary, and neither produces a biological conclusion without identification, quantification and context. The practical capability is metabolomics literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “The metabolome contains many small molecules” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A reported metabolome is the portion detectable by a chosen workflow.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Fasting state, collection tube, delay, temperature, centrifugation, storage and freeze-thaw cycles can alter samples. Randomisation and metadata help separate biology from handling.

Untargeted analysis is a discovery workflow with many features and statistical choices. Targeted assays answer narrower questions with standards, calibration and defined performance. The practical capability is pre-analytical reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Pre-analytics begin before the instrument” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A sophisticated instrument cannot recover molecules changed before analysis.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Groups should be comparable for relevant age, sex, diet, medicines, batch and collection conditions. Repeated measures or matching can improve inference when the question and analysis are planned in advance.

A biomarker must survive more than discovery. Analytical validity, biological reproducibility, independent validation and usefulness for a defined decision are separate milestones. The practical capability is metabolomics study design. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Study design controls confounding” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A large sample does not automatically remove systematic confounding.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Liquid chromatography moves compounds through a column where chemistry affects retention. Gradient, column, temperature and matrix influence separation and must remain documented.

The metabolome is dynamic. Food, medicines, activity, microbes, time of day and sample handling can shift measured molecules, so study design and pre-analytical records are essential. The practical capability is chromatography literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Chromatography separates mixtures in time” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A retention time alone does not prove molecular identity.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Ionised molecules and fragments are separated by mass-to-charge ratio. Ion source, adducts, collision energy, resolution and calibration shape the feature table.

Mass spectrometry and NMR observe chemistry through different physical signals. Their strengths are complementary, and neither produces a biological conclusion without identification, quantification and context. The practical capability is mass-spectrometry reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Mass spectrometry measures mass-to-charge signals” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: One signal may represent an adduct, isotope or fragment rather than a unique metabolite.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Nuclear magnetic resonance detects resonant behaviour of nuclei in a magnetic field. Chemical shifts and coupling support structural and quantitative interpretation with suitable references.

Untargeted analysis is a discovery workflow with many features and statistical choices. Targeted assays answer narrower questions with standards, calibration and defined performance. The practical capability is NMR literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “NMR reads magnetic environments” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: Overlapping peaks can limit identification in complex mixtures.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Untargeted metabolomics detects many reproducible features, then aligns peaks, filters noise and tests patterns. Discovery power comes with multiple comparisons and uncertain identities.

A biomarker must survive more than discovery. Analytical validity, biological reproducibility, independent validation and usefulness for a defined decision are separate milestones. The practical capability is untargeted-analysis reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Untargeted workflows search broadly” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: An untargeted feature is not a validated biomarker.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Targeted methods use standards, transitions and calibration to quantify predefined metabolites. Selectivity, range, precision, recovery and stability are evaluated for the intended matrix.

The metabolome is dynamic. Food, medicines, activity, microbes, time of day and sample handling can shift measured molecules, so study design and pre-analytical records are essential. The practical capability is targeted-assay literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Targeted assays answer narrower questions” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A targeted panel cannot discover compounds outside its list.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Blanks detect contamination, pooled samples track repeatability, standards monitor response and randomised batches reduce confounding. Control charts can flag maintenance or correction needs.

Mass spectrometry and NMR observe chemistry through different physical signals. Their strengths are complementary, and neither produces a biological conclusion without identification, quantification and context. The practical capability is analytical-quality control. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Quality-control samples reveal drift” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: Post-hoc correction should not hide a failing analytical run.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Dilution, sample amount, internal standards and batch effects can influence intensities. The chosen normalisation assumes what should remain stable and must be tested with sensitivity analyses.

Untargeted analysis is a discovery workflow with many features and statistical choices. Targeted assays answer narrower questions with standards, calibration and defined performance. The practical capability is data-normalisation reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Normalisation changes the comparison” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A normalised value is not an absolute concentration unless calibrated.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Accurate mass, isotope pattern, fragments, retention and authentic standards provide increasing confidence. Reporting levels prevent tentative annotations from appearing definitive.

A biomarker must survive more than discovery. Analytical validity, biological reproducibility, independent validation and usefulness for a defined decision are separate milestones. The practical capability is metabolite-identification literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Identification requires converging evidence” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A database match by mass alone is not confirmed identity.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Lipids vary by head group, chain length, saturation and bond position. Extraction, ionisation and fragmentation favour different classes, so panels need explicit coverage.

The metabolome is dynamic. Food, medicines, activity, microbes, time of day and sample handling can shift measured molecules, so study design and pre-analytical records are essential. The practical capability is lipidomics reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Lipidomics faces structural complexity” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A total lipid label can hide biologically different species.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Genomics, transcriptomics, proteomics, metabolomics and microbiome data have different scales and missingness. Integration should start with a biological question and preserve each layer’s quality controls.

Mass spectrometry and NMR observe chemistry through different physical signals. Their strengths are complementary, and neither produces a biological conclusion without identification, quantification and context. The practical capability is multi-omics literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Multi-omics integration aligns different layers” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: More omics layers do not guarantee a clearer mechanism.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Feature selection, cross-validation, held-out data and multiple-testing correction reduce false discoveries. Batch and cohort structure must stay separated from outcome labels.

Untargeted analysis is a discovery workflow with many features and statistical choices. Targeted assays answer narrower questions with standards, calibration and defined performance. The practical capability is metabolomics data science. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Statistics should protect against overfitting” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A visually separated score plot is not independent validation.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

A candidate needs analytical performance, biological replication, independent cohorts, thresholds and evidence that it improves a defined decision. Mechanistic studies can explain why it tracks an outcome.

A biomarker must survive more than discovery. Analytical validity, biological reproducibility, independent validation and usefulness for a defined decision are separate milestones. The practical capability is biomarker-validation reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Biomarkers progress through validation stages” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: Association does not establish diagnostic or therapeutic usefulness.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Archive consent and cohort, specimen, collection and storage, preparation, column and instrument, standards, QC, raw files, processing code, identities, statistics, uncertainty and decision stage.

The metabolome is dynamic. Food, medicines, activity, microbes, time of day and sample handling can shift measured molecules, so study design and pre-analytical records are essential. The practical capability is metabolomics career preparation. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. NTU SPC and project owners control current services, access and released findings; clinicians, laboratories and responsible authorities control diagnostic or patient-care use.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Keep a metabolomics evidence receipt” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: Only current qualified laboratories and responsible authorities can approve diagnostic use.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter