Surface predictive and prognostic subgroup evidence with population modifiers and conflicting-effect flags. Motif does not enroll patients or run clinical decision support.
The pipeline
Search, extract, inspect modifiers, and detect flips. This is literature evidence for trial design, not patient enrollment. For enrichment design context, read our blog on patient stratification in clinical trials to learn more.
Plain-language input
Ask which biomarkers predict treatment benefit or prognosis in a disease context. Motif searches PubMed, PMC, and Europe PMC, screens relevance on title and abstract, and extracts associations with predictive and prognostic predicates. For checkpoint inhibitor predictors, read our blog on immunotherapy biomarkers to learn more.
Hi Alex, what would you like to explore?
1. PD-L1 TPS ≥50% predicts improved PFS on pembrolizumab vs chemotherapy in untreated NSCLC (HR=0.50, 95% CI: 0.37–0.68, RCT n=305)
↓ Supporting Evidence (2)
2. Tumor mutational burden-high is prognostic for shorter OS in males but not females with advanced melanoma (interaction p=0.02, n=810)
↓ Supporting Evidence (2)
3. HER2 amplification predicts trastuzumab benefit in metastatic breast cancer (HR=0.67, comparator present, validation cohort n=287)
↓ Supporting Evidence (3)
Association library
Motif reads full text and emits association sentences with effect sizes, study design, and GRADE-adapted evidence tiers. Each claim links to a PMID. Discovery and validation cohorts appear as separate associations when papers report independent replication.
Subgroup context
Open any association to see biological context modifiers with stratum sample sizes, comparator presence, and interaction p-values. This is the evidence layer trial designers need before writing enrichment criteria, not an automated patient-selection engine.
Association detail
PD-L1 TPS ≥50% predicts improved PFS on pembrolizumab vs chemotherapy in untreated NSCLC
Actor
PD-L1 (protein)
Predicate
predictive
Target
pembrolizumab (therapeutic)
Biological context modifiers
Stratification flip detection
2 opposing effects across strata
TMB-high → pembrolizumab response
Modifier category: Demographic
Male
predictive benefit
Female
no significant benefit
Severity: Moderate
EGFR → MAPK pathway activation
Modifier category: Stage
Stage III
Higher expression
Stage IV
Lower expression
Severity: Low
Conflicting subgroups
Motif compares biomarker effects across subgroups and flags when the same marker shows opposing results in different strata, for example benefit in one sex but not another. Broader conflicts across studies are surfaced separately so you can see where published evidence disagrees. Read our blog on personalized medicine biomarker analysis to learn more about broader precision-medicine context.
What Motif does and does not do for trial design
No. Motif surfaces published biomarker evidence with subgroup modifiers and PMIDs. It does not access patient records, run clinical decision support, or assign patients to treatment arms. Your team uses the extracted evidence to design enrichment criteria and inclusion rules.
Predictive associations link a biomarker to differential treatment benefit. They require a therapeutic target and often a comparator arm. Prognostic associations forecast outcome in the studied population regardless of treatment contrast. Motif extracts both with distinct predicates and surfaces comparator and interaction fields where reported.
Modifiers describe the subgroup context of an association, for example stage, molecular subtype, demographic group, or discovery vs validation cohort. They appear in the association detail panel with stratum sample sizes and category labels drawn from Motif's 16 modifier categories.
Motif compares associations for the same biomarker and outcome across different subgroup values within one modifier type, such as sex, stage, or line of therapy. When effects point in opposite directions, Motif flags the pair with a severity level based on evidence certainty. Review the cited papers before changing enrollment criteria.
Yes. The Association Library supports filters for enrollment-defined stratum, comparator presence, maximum interaction p-value, minimum evidence certainty, and biological context. These filters help you focus on evidence where the paper explicitly selected patients by the biomarker.
Evidence quality depends on what authors report in abstracts and full text. Subgroup analyses are often underpowered or exploratory. Motif does not replace biostatistical review of your protocol. Scope depends on your plan: Starter searches up to 5 papers per query, Pro up to 40. Verify key subgroup claims in the source papers.
Join researchers using Motif to find predictive and prognostic biomarker evidence from published literature.