Skip to main content

Grant & Manuscript Pipeline

Claude Code for the full academic production cycle - from BCO-DMO documentation to reviewer responses

25 min
10 min read

Grant & Manuscript Pipeline

Research isn't just analysis. Here's how Claude Code fits into the full academic production cycle.

Data Documentation for Repositories

BCO-DMO Submission Package

NSF requires data archiving. Make it less painful.

Bash
claude "I need to prepare my dataset for BCO-DMO submission.
Read data/coral_survey_final.csv and create a complete submission package:
1. DATA DICTIONARY
- All column names, types, units
- Allowed values/ranges
- Missing value codes used
- Link each variable to collection method (I'll fill details)
2. ABSTRACT (250 words)
- What data this is
- How it was collected
- Geographic/temporal scope
- Potential uses
3. METHODS DOCUMENT
- Sampling design
- Collection protocols
- Quality control procedures
- Processing steps applied
- Known limitations
4. METADATA FILE (ISO 19115 format)
- Following BCO-DMO template structure
5. README.txt
- File inventory
- How to cite
- Contact information
- Related publications
Export each as separate files matching BCO-DMO naming conventions."

Zenodo/Figshare Documentation

Bash
claude "Create documentation for archiving this analysis on Zenodo:
1. README.md covering:
- Project description
- File inventory with descriptions
- How to reproduce the analysis
- Citation instructions
- License information (suggest appropriate one)
2. CITATION.cff file for proper citation tracking
3. Suggested keywords and description for the Zenodo upload form
Match the style of highly-cited data repositories in ecology."

Supplementary Materials

Generating Supplement Structure

Bash
claude "Based on our main analysis, create a supplementary materials document:
STRUCTURE NEEDED:
- Table S1: Full model results (not truncated like Table 2)
- Table S2: Sensitivity analysis results
- Table S3: Site characteristics
- Figure S1: Diagnostic plots for main model
- Figure S2: Map of study sites
- Methods S1: Detailed statistical methodology
For each item:
1. Generate the content from our analysis results
2. Add appropriate captions
3. Cross-reference to main text
Format for Nature Ecology & Evolution style.
Export as both Word (.docx) and LaTeX (.tex)."

Converting Between Formats

Bash
claude "Convert this supplementary material:
- FROM: supplementary.tex (LaTeX)
- TO: supplementary.docx (Word)
Preserve:
- Table formatting
- Figure references
- Equation numbering
- Citation style
My co-author needs Word and won't learn LaTeX."

Reviewer Response Management

Organizing Comments

Bash
claude "Parse this reviewer response document and create a structured response template:
REVIEWER COMMENTS:
[paste the full reviewer letter]
CREATE:
1. A numbered list of every distinct comment/request
2. Categorize each as:
- Minor: Quick fix
- Moderate: Some work needed
- Major: Significant revision
- Disagree: We should push back
3. For each comment, create response skeleton:
- Our response: [to fill]
- Changes made: [to fill]
- Location in manuscript: [to fill]
4. Track which analyses/figures need updating
Export as Word table I can fill in with co-authors."

Drafting Responses

Bash
claude "Draft a response to this reviewer comment:
COMMENT: 'The authors should justify their choice of mixed-effects
model over a simpler repeated-measures ANOVA. The sample size
(n=47) may be insufficient for the complexity of their model.'
CONTEXT:
- We chose mixed-effects because we have unbalanced data
- Random slopes are needed due to site-specific variation
- We did run power analysis during design phase
Draft a response that:
1. Acknowledges the valid concern
2. Explains our rationale clearly
3. Cites methodological literature supporting our choice
4. Describes any additional analyses we'll add to address this
5. Is diplomatic but defends our approach
Tone: Respectful but confident. Don't be obsequious."

Response Letter Assembly

Bash
claude "Compile our individual responses into a formatted response letter:
Include:
1. Cover letter to editor (brief, professional)
2. Point-by-point responses organized by reviewer
3. Summary of major changes
4. Track changes reference guide
Format conventions:
- Reviewer comments in italic
- Our responses in regular text
- Changed manuscript text in quotes with line numbers
- New analyses/figures highlighted
Generate both Word and PDF versions."

Manuscript Preparation

Methods Section Drafting

Bash
claude "Based on my analysis code in analysis/main.R, draft a Methods section:
REQUIREMENTS:
- Sufficient detail for reproducibility
- Statistical terminology appropriate for [journal]
- Report all test statistics, effect sizes, confidence intervals
- Explain any non-standard approaches
- Reference software versions
STRUCTURE:
1. Study design and sampling
2. Data collection procedures
3. Statistical analysis
- Data cleaning and preparation
- Primary analysis approach
- Secondary/sensitivity analyses
4. Software and packages used
LENGTH: ~500 words for PLOS ONE format
I'll add the domain-specific details; you handle the statistical reporting."

Results Section Support

Bash
claude "Generate a Results section skeleton from my analysis outputs:
For each significant finding:
1. State the finding in plain language
2. Report statistics properly: F(df1, df2) = X.XX, p = .XXX, η² = .XX
3. Reference the appropriate table/figure
4. Note effect size interpretation
For non-significant findings:
1. Report the statistics
2. Note the power/sample size considerations
DO NOT interpret—just report. I'll add the interpretation."

Figure Captions

Bash
claude "Write publication-ready figure captions for these figures:
Figure 1: [describe what it shows]
Figure 2: [describe what it shows]
Figure 3: [describe what it shows]
Each caption should:
1. Start with a single-sentence summary
2. Describe what is being plotted (axes, colors, symbols)
3. Explain any statistical annotations
4. Define abbreviations on first use
5. Note sample sizes
Style: Nature Ecology & Evolution
Maximum: 350 words per caption"

Grant Writing Support

Specific Aims Iteration

Bash
claude "I'm writing an NSF proposal. Here are my three specific aims:
[paste current aims]
Help me strengthen them:
1. Are they logically connected?
2. Is each aim achievable in the timeline?
3. Do they build on each other or are they independent?
4. Suggest how to make the impact statement stronger
5. Flag any red flags a reviewer might raise
Be critical—I'd rather fix problems now than get triaged."

Data Management Plan

Bash
claude "Generate a Data Management Plan for my NSF proposal:
PROJECT: [brief description]
DATA TYPES:
- Field observations (CSV, ~10GB)
- Photographs (JPEG, ~50GB)
- Derived data products (NetCDF, ~5GB)
- Analysis code (R scripts)
REQUIREMENTS TO ADDRESS:
1. What data will be generated
2. Data formats and standards
3. Data storage during project
4. Data sharing and access
5. Long-term preservation
6. Roles and responsibilities
Follow NSF BIO requirements.
Include specific repository names (BCO-DMO, Zenodo, GitHub).
Keep to 2 pages."

Budget Justification Text

Bash
claude "Draft budget justification text for these line items:
1. Graduate student (2 years, 50% time): $XX,XXX
2. Undergraduate assistants (3 students, summer): $X,XXX
3. Field equipment (list attached): $XX,XXX
4. Travel (2 field seasons + 1 conference/year): $XX,XXX
5. Publication costs: $X,XXX
For each item:
- Explain what they'll do
- Why this amount/duration is appropriate
- How it connects to the proposed work
Tone: Clear, justified, not defensive."

LaTeX Workflow Integration

Compiling Multi-Author Documents

Bash
claude "Help me set up a collaborative LaTeX project:
CREATE:
1. Main.tex (master document that includes chapters)
2. Chapters/ directory structure for sections
3. A preamble.tex with common packages and commands
4. A Makefile for easy compilation
5. .gitignore for LaTeX projects
INCLUDE:
- Natbib setup for [citation style]
- Cross-referencing with cleveref
- Draft mode with line numbers
- Clean compilation without warnings
My co-authors will edit separate chapter files."

Converting Analysis Output to LaTeX Tables

Bash
claude "Convert this regression output to a publication-ready LaTeX table:
[paste R/Python output or CSV]
REQUIREMENTS:
- Three-line table style (no vertical lines)
- Proper alignment (left for text, right for numbers)
- Significance stars with footnote explaining
- Standard errors in parentheses
- Model fit statistics at bottom
- Booktabs package formatting
Match the style of tables in Journal of Ecology."

BibTeX Management

Bash
claude "Clean up my BibTeX file:
Read references.bib and:
1. Fix inconsistent formatting
2. Fill in missing fields where you can infer them
3. Flag entries that seem incomplete
4. Remove duplicate entries (keep most complete version)
5. Standardize journal name abbreviations
6. Check DOIs are properly formatted
Also create a 'needed_refs.txt' listing any citations in my .tex files
that aren't in the .bib file."

Multi-Format Output

One Analysis, Many Outputs

Bash
claude "From this analysis, generate outputs for different audiences:
1. TECHNICAL (for co-authors):
- Full statistical output
- All diagnostics
- Code and data references
2. MANUSCRIPT (for submission):
- Cleaned tables (Tables 1-3)
- Publication figures (Figures 1-4)
- Results text snippets
3. PRESENTATION (for conference):
- Key findings as bullet points
- Simplified figures (higher contrast, bigger fonts)
- Talking points for each slide
4. SUMMARY (for stakeholders):
- Plain language summary (1 paragraph)
- Key numbers without statistics
- Implications for practice
Export each to appropriate formats."

Workflow Templates

Paper Submission Checklist

Bash
claude "Create a submission checklist for [journal]:
PRE-SUBMISSION:
- [ ] All co-authors have approved final version
- [ ] All figures meet journal specifications
- [ ] References formatted correctly
- [ ] Word/page count within limits
- [ ] Cover letter drafted
- [ ] Suggested reviewers identified
- [ ] Competing interests declared
REQUIRED FILES:
- [ ] Main manuscript (format: ___)
- [ ] Figures (format: ___, resolution: ___)
- [ ] Tables (format: ___)
- [ ] Supplementary materials
- [ ] Data availability statement
- [ ] Author contributions (CRediT format)
METADATA NEEDED:
- [ ] All author ORCIDs
- [ ] Keywords (n = ___)
- [ ] Running title (max ___ characters)
- [ ] Corresponding author details
Pull specific requirements from [journal] guidelines."

Common Pain Points Solved

"Track Changes Is a Mess"

Bash
claude "Compare these two manuscript versions and create a clean summary:
- v2_from_advisor.docx
- v2_from_coauthor.docx
Show me:
1. Changes they both made (keep these)
2. Conflicting changes (I need to decide)
3. Changes only one person made (review needed)
Don't generate a merged file—just help me see what needs reconciling."

"Reference Numbers Shifted"

Bash
claude "My figure/table numbers changed during revision.
Search the manuscript and update all references:
OLD -> NEW mapping:
- Figure 2 -> Figure 3
- Figure 3 -> Figure 2
- Table 1 -> Table 2
- Table S1 -> Table S2
Find and update all occurrences in main text and supplementary materials."

"The Journal Wants Different Figure Sizes"

Bash
claude "Regenerate all figures at these specifications:
- Single column: 89mm wide
- Double column: 183mm wide
- Full page: 183mm x 247mm
- Resolution: 300 DPI for photos, 600 DPI for line art
- Format: TIFF with LZW compression
- Fonts: Arial, minimum 8pt
Current figures are in figures/. Match content exactly, just resize."

Next Steps