Q (haiku): What is the median survival time in the Pediatric Neuroblastoma study from TARGET?

▶ read_guide
{
 "uri": "cbioportal://statistical-tests-guide"
}

◀ result
Statistical Test Selection Guide
================================

Purpose
-------
This guide ensures the correct statistical test is selected before performing any group comparison, matching cBioPortal's own Group Comparison defaults.

HARD RULES — NEVER FABRICATE A STATISTIC
----------------------------------------
ClickHouse cannot run statistical tests. The agent therefore must NEVER produce a derived statistic that is not a literal column value from a SQL result. Specifically:

1. **Never invent a p-value.** Not "p < 0.001", not "p ≈ 0.05", not any p-value. If the user asks "what is the p-value?", the answer is *"I can't compute that — here is the 2x2 contingency table (or group statistics). Run it in cBioPortal's Group Comparison tab, in R with `fisher.test(...)` / `wilcox.test(...)`, or in Python with `scipy.stats.fisher_exact(...)` / `mannwhitneyu(...)`."*
2. **Never claim mutual exclusivity (or co-occurrence) from a contingency table alone.** A 2x2 table is not a test. The shape "altered/not altered × group A/group B" needs Fisher's exact + a defined direction (odds ratio < 1 with significant p). Without that test, the agent presents the table and stops. Descriptive phrasing is also a claim: "largely/mostly mutually exclusive", "rarely co-occur", "tend to co-occur" are forbidden without the test — point to cBioPortal's Mutual Exclusivity tab instead.
3. **Never report a "median" that came from `AVG(...)` or any non-median aggregate.** "Median" and "mean" are different statistics; for skewed clinical distributions (especially survival) they differ substantially. Use ClickHouse's `quantile(0.5)(...)` for actual median, and label arithmetic averages as "mean", never "median".
4. **Never report a hazard ratio, odds ratio, risk ratio, or relative risk** that wasn't computed by an external tool. These require regression / model fitting that ClickHouse does not do.
5. **Never report median overall survival from `AVG(OS_MONTHS)` or even `quantile(0.5)(OS_MONTHS)`.** Median OS requires Kaplan-Meier estimation, which handles censoring (`OS_STATUS = 0:LIVING` means the event hasn't happened yet). Naive medians/means over `OS_MONTHS` ignore censoring and are systematically wrong. The correct handoff: return the raw `(OS_MONTHS, OS_STATUS)` pairs (or descriptive counts: N events, N censored, follow-up range) and tell the user to run KM in R (`survival::survfit`) or Python (`lifelines.KaplanMeierFitter`), or use cBioPortal's Survival comparison.

If asked to do any of (1)–(5), respond with the appropriate handoff template from the "Approved Response Templates" section below — do not produce the number.

Ambiguous-Term Clarification (before any analysis)
--------------------------------------------------
If the user's question uses an ambiguous term, ask before computing anything. Examples:

- **"aggressive"** → could mean shorter overall survival, higher metastasis rate, higher grade/stage, higher TMB, or specific molecular features. Ask which definit … (9984 more chars)

▶ list_studies
{
 "search": "TARGET Neuroblastoma",
 "verbose": true
}

◀ result
{"result":[{"cancer_study_identifier":"nbl_target_gdc","name":"Neuroblastoma (TARGET GDC, 2025)","description":"TARGET Neuroblastoma. Source data from <A HREF=\"https://gdc.cancer.gov\">NCI GDC</A> and generated in Aug 2025 using <A HREF=\"https://cda.readthedocs.io/en/latest/\">Cancer Data Aggregator</A>.","type_of_cancer_id":"nbl","sample_count":533,"has_guide":false,"url":"https://www.cbioportal.org/study/summary?id=nbl_target_gdc"},{"cancer_study_identifier":"nbl_target_2018_pub","name":"Pediatric Neuroblastoma (TARGET, 2018)","description":"Whole genome or whole exome sequencing on 1089 NBL samples. Comprehensive profiling of 1089 NBL samples. <p>TARGET data is intended exclusively for biomedical research using pediatric data (i.e., the research objectives cannot be accomplished using data from adults) that focus on the development of more effective treatments, diagnostic tests, or prognostic markers for childhood cancers. Moreover, TARGET data can be used for research relevant to the biology, causes, treatment and late complications of treatment of pediatric cancers, but is not intended for the sole purposes of methods and/or tool development (please see <a href=\"https://ocg.cancer.gov/programs/target/using-target-data\">Using TARGET Data</a> section of the OCG website). If you are interested in using TARGET data for publication or other research purposes, you must follow the <a href=\"https://ocg.cancer.gov/programs/target/target-publication-guidelines\">TARGET Publication Guidelines</a>.</p>","type_of_cancer_id":"nbl","sample_count":1089,"has_guide":false,"url":"https://www.cbioportal.org/study/summary?id=nbl_target_2018_pub"},{"cancer_study_identifier":"nbl_msk_2023","name":"Pediatric Neuroblastoma (MSK, Nat Genet 2023)","description":"Targeted sequencing of 223 pediatric neuroblastoma tumor/normal pairs via MSK-IMPACT.","type_of_cancer_id":"nbl","sample_count":223,"has_guide":false,"url":"https://www.cbioportal.org/study/summary?id=nbl_msk_2023"}]}

▶ clickhouse_run_select_query
{
 "query": "
SELECT
  COUNT(DISTINCT patient_unique_id) AS total_patients,
  COUNT(DISTINCT CASE WHEN os_status = '1:DECEASED' THEN patient_unique_id END) AS n_events,
  COUNT(DISTINCT CASE WHEN os_status = '0:LIVING' THEN patient_unique_id END) AS n_censored,
  MIN(os_months) AS min_followup_months,
  MAX(os_months) AS max_followup_months,
  ROUND(AVG(os_months), 2) AS mean_followup_months
FROM clinical_data_derived
WHERE cancer_study_identifier = 'nbl_target_2018_pub'
  AND os_months IS NOT NULL
  AND os_status IS NOT NULL
",
 "max_rows": 10
}

◀ result
{"error_message":"Query execution failed: HTTPDriver for https://dl96orhu96.us-east-1.aws.clickhouse.cloud:8443 received ClickHouse error code 47\n Code: 47. DB::Exception: Unknown expression or function identifier `os_status` in scope SELECT COUNTDistinct(patient_unique_id) AS total_patients, COUNTDistinct(multiIf(os_status = '1:DECEASED', patient_unique_id, NULL)) AS n_events, COUNTDistinct(multiIf(os_status = '0:LIVING', patient_unique_id, NULL)) AS n_censored, MIN(os_months) AS min_followup_months, MAX(os_months) AS max_followup_months, ROUND(AVG(os_months), 2) AS mean_followup_months FROM clinical_data_derived WHERE (cancer_study_identifier = 'nbl_target_2018_pub') AND (os_months IS NOT NULL) AND (os_status IS NOT NULL). (UNKNOWN_IDENTIFIER) (version 26.4.1.2359 (official build))\n"}

▶ clickhouse_list_table_columns
{
 "table": "clinical_data_derived"
}

◀ result
{"columns":[{"name":"internal_id","type":"Int32"},{"name":"sample_unique_id","type":"String","comment":"Globally unique sample ID: cancer_study_identifier + \"_\" + sample.stable_id. Empty for patient-level attributes. Use this for sample filtering and joins."},{"name":"patient_unique_id","type":"String","comment":"Globally unique patient ID: cancer_study_identifier + \"_\" + patient.stable_id. Present for both sample and patient-level attributes."},{"name":"attribute_name","type":"LowCardinality(String)","comment":"Clinical attribute name (e.g., SAMPLE_TYPE, CANCER_TYPE, AGE, OS_MONTHS). Use with attribute_value for filtering. AGE may be floored or capped for de-identification (e.g. all children recorded as 18, or everyone 89+ recorded as 89 or 90): before age statistics check for a pile-up at the min/max, and if present compute age from DAYS_TO_BIRTH (-days / 365.25)."},{"name":"attribute_value","type":"String","comment":"Value of the clinical attribute (String). For SAMPLE_TYPE: Primary, Metastasis, Local Recurrence, Unknown. Missing values are empty strings, so use toFloat64OrNull(attribute_value) for numeric comparisons — CAST fails on them."},{"name":"cancer_study_identifier","type":"LowCardinality(String)"},{"name":"type","type":"LowCardinality(String)","comment":"Data level: \"sample\" for sample-level attributes (e.g., SAMPLE_TYPE), \"patient\" for patient-level attributes (e.g., AGE, OS_MONTHS)."}]}

▶ clickhouse_run_select_query
{
 "query": "
SELECT
  COUNT(DISTINCT patient_unique_id) AS total_patients,
  COUNT(DISTINCT CASE WHEN 
    attribute_name = 'OS_STATUS' AND attribute_value = '1:DECEASED' 
    THEN patient_unique_id END) AS n_deceased_events,
  COUNT(DISTINCT CASE WHEN 
    attribute_name = 'OS_STATUS' AND attribute_value = '0:LIVING' 
    THEN patient_unique_id END) AS n_censored_living,
  MIN(CASE WHEN attribute_name = 'OS_MONTHS' THEN toFloat64OrNull(attribute_value) END) AS min_followup_months,
  MAX(CASE WHEN attribute_name = 'OS_MONTHS' THEN toFloat64OrNull(attribute_value) END) AS max_followup_months
FROM clinical_data_derived
WHERE cancer_study_identifier = 'nbl_target_2018_pub'
  AND attribute_name IN ('OS_MONTHS', 'OS_STATUS')
GROUP BY cancer_study_identifier
",
 "max_rows": 10
}

◀ result
{"rows":[{"total_patients":1076,"n_deceased_events":397,"n_censored_living":675,"min_followup_months":1,"max_followup_months":183}]}

═ answer (success)
## Median Overall Survival — Pediatric Neuroblastoma (TARGET, 2018)

**Important Note:** Median overall survival **cannot be computed from raw data alone** because survival data is *censored* — patients still living at last follow-up have not yet experienced the event. A naive median or mean ignores censoring and produces a systematically wrong answer.

### Cohort Summary
- **Total patients with survival data:** 1,076
- **N events (deceased):** 397
- **N censored (alive at last follow-up):** 675
- **Follow-up range:** 1–183 months

### How to Obtain the Correct Median OS

**Median overall survival requires Kaplan-Meier estimation**, which accounts for censoring. Here are your options:

1. **cBioPortal's Survival comparison page** (recommended — built-in, no tools needed):
   - [View Pediatric Neuroblastoma (TARGET, 2018) study](https://www.cbioportal.org/study/summary?id=nbl_target_2018_pub)
   - On the study page, click **"Survival"** to see the cohort-wide KM curve with median OS marked

2. **R** (`survival` package):
   ```R
   library(survival)
   fit <- survfit(Surv(OS_MONTHS, OS_STATUS=="1:DECEASED") ~ 1, data=your_data)
   summary(fit)  # median OS is in the output
   ```

3. **Python** (`lifelines` library):
   ```python
   from lifelines import KaplanMeierFitter
   kmf = KaplanMeierFitter()
   kmf.fit(T=os_months, E=(os_status=="1:DECEASED"), label="NBL")
   print(kmf.median_survival_time_)
   ```

Given that only 37% of patients (397/1,076) experienced the event during follow-up, **the Kaplan-Meier median is likely not reached** — meaning more than half of patients were still alive at last follow-up. cBioPortal's KM curve will show this clearly.