Sequencing analysis
Choose a microRNA source study and test its claim
Read the task instructions and deliverables
# Choose a microRNA source study and test its claim Your group can study either MIR4664 or MIR3175 in MCF7 cells. Choose the stronger lead for a follow-up test of a separate RNA transcription source, using evidence repeated across the four supplied libraries. This is a research priority, not a claim that a separate source is proved. For the chosen microRNA, make two separate decisions: does the evidence identify RNA that starts being made separately, and does it establish increased RNA production under the recorded E2 condition? The second decision does not require a separate source. Use quantitative analysis of all four FASTQs and the original records to decide which parts of the biological claim the evidence supports. Try to disprove your priority. Consider the strongest competing explanation, and identify an observation that could distinguish it from your favored explanation. Choose your own measurements; no particular interval or summary is required. PRO-seq measures RNA while it is being made. Submit only findings.json with the three generated final-result fields. Supporting analysis is required to reach the decisions, but the final decisions alone are graded; no prose or particular supporting measurement earns a point. The FASTQs are uniform subsets of 5,000,000 complete records per original library, preserving both sequencing lanes. If reporting read counts in your working analysis, count primary alignments and count separate read records at the same position separately. State units and which records a reported measurement covers. Use only supplied evidence and at most four compute threads.
How was it graded?
Code compares submitted answer fields with reference values. Written reasoning is not graded.
Sum of passed criterion weights divided by total criterion weights.
LLM judge: None — deterministic grading
Task files
Choose an evidence file.
Grading criteria
A one-item list giving the gene symbol or supplied gene ID of the microRNA to study first, as defined in the question. This recommends a follow-up; it does not assert that separate transcription is established.
Earns credit
The priority field identifies MIR3175. Its stronger repeated spatial contrast supports this follow-up priority. The shared runtime checks only the final selection; supporting prose is not graded.
Does not earn credit
The final result is missing, has an invalid type, or fails the shared comparison. Incorrect target, label direction, population, priority or source claim does not earn this criterion.
Accepted alternatives
Order does not affect the set comparison. Repeated members, including aliases of the same member, fail the check. Letter case is ignored. Leading and trailing whitespace is ignored. Accepted aliases: ensg00000284324.1 → MIR3175; ensg00000284324 → MIR3175; hsa-mir-3175 → MIR3175; mir-3175 → MIR3175.
Weight: 4
Input evidence
All 5,000,000 records · Exact library input; no processed measurements supplied.
Reference evidence
reference-primary · /by_threshold/0/panels; compare both sites over 100, 250, 500 and 1000 bp padding, all four libraries, three signal summaries. Larger windows are a sensitivity challenge, not a required diagnostic.
MIR3175 is the clearer localized nascent lead; that ranking does not establish separate initiation. Moderate region widths and all libraries support it, while excessive widths can absorb other signal.
reference-prefix · /by_threshold/0/panels; compare both sites over 100, 250, 500 and 1000 bp padding, all four libraries, three signal summaries. Larger windows are a sensitivity challenge, not a required diagnostic.
Independently rerun adapter-prefix variant on all four exact final FASTQs; endpoint definition is the same.
reference-historical-local · /by_threshold/0/panels; compare both sites over 100, 250, 500 and 1000 bp padding, all four libraries, three signal summaries. Larger windows are a sensitivity challenge, not a required diagnostic.
Retained local-alignment measurements verified against exact FASTQ hashes, receipts and span histograms, then recomputed using the final endpoint definition.
Recorded result / GPT 6.1 sol
Failed. priority: matched 0 of 1 reference members. Missing: MIR3175. Extra: mir4664.
True if the supplied evidence identifies a separately initiated RNA source for the priority microRNA; false if this is not established. This asks where RNA starts being made, not whether the microRNA exists.
Earns credit
The source_established field is false. Separate initiation is not established. Compatible containing-gene transcription and local accumulation remain unresolved sources; spatial concentration does not identify a start site. The shared runtime checks only the final boolean.
Does not earn credit
The boolean is true, missing, null, or has the wrong type.
Accepted alternatives
Any scientifically valid evidence may support this final decision. False means unestablished, not absent. No prose or specific measurement is graded.
Weight: 1
Input evidence
All 5,000,000 records · Exact library input; no processed measurements supplied.
Reference evidence
scientific-review · Established claims; Competing sources; occupancy versus production flux
Separate initiation is not established. Compatible containing-gene transcription and local accumulation remain unresolved sources; spatial concentration does not identify a start site.
reference-primary · /by_threshold/0/ratios; /by_threshold/0/panels; /annotation_overlap; /orientation
Exact-input physical measurements underlying the source interpretation and its counterevidence.
Recorded result / GPT 6.1 sol
Passed. Submitted source_established = false; required the JSON boolean false.
True if the supplied evidence establishes increased RNA production at the priority microRNA locus under the recorded E2 condition; false if this is not established. A separate RNA source is not required. False does not mean an increase is biologically impossible.
Earns credit
The increase_established field is false. Increased production is not established. A descriptive increase in short-feature occupancy coexists with weak broader changes, possible pausing or elongation differences, and unresolved treatment/replicate context. Occupancy alone does not establish RNA production flux. The shared runtime checks only the final boolean.
Does not earn credit
The boolean is true, missing, null, or has the wrong type.
Accepted alternatives
Any scientifically valid evidence may support this final decision. False means unestablished, not absent. No prose or specific measurement is graded.
Weight: 1
Input evidence
All 5,000,000 records · Exact library input; no processed measurements supplied.
Reference evidence
scientific-review · Established claims; Competing sources; occupancy versus production flux
Increased production is not established. A descriptive increase in short-feature occupancy coexists with weak broader changes, possible pausing or elongation differences, and unresolved treatment/replicate context. Occupancy alone does not establish RNA production flux.
reference-primary · /by_threshold/0/ratios; /by_threshold/0/panels; /annotation_overlap; /orientation
Exact-input physical measurements underlying the source interpretation and its counterevidence.
Recorded result / GPT 6.1 sol
Passed. Submitted increase_established = false; required the JSON boolean false.