Considering setting up an AI Agent for trt research and looking for input

Phil Goodman · Apr 22, 2026

Been thinking about this for a while and might actually go through with the project. As someone in the tech field who has been interested in health and research for decades I’ve started to consider the combing the two now that tech has reached it’s current state(and only getting better). These seem to be some of the best use cases for AI agents these days(combing through TONS of research and other data and building models) so I’m interested in applying it to the world of trt. But obviously you run the risk of the dreaded garbage in/garbage out results… so I’d be interested to get feedback on what type of stuff should be included and how it should be modeled. For one, I want to reduce risk of my bias creeping in. Also, there are lots of smart people here so I’m sure I can get plenty of good suggestions and perspectives that I wouldn’t consider on my own. To note though, while I use it regularly for lots of analysis and am comfortable with the technology(and tech in general), there will be a learning curve starting out… though a lot of that can be eliminated by proper use of the AI itself. For example, below I’ll share some of the guidance so far. Seems like a good starting point, but again there will probably be some bumps and tweaking along the way.

Literature review & critique — AI is excellent at this right now. Claude, GPT-4, and others can ingest hundreds of papers, identify methodological weaknesses (small n, short duration, lack of controls, industry funding bias, etc.), and synthesize findings.
“Running millions of simulations” — this is the part to think carefully about. AI doesn’t run simulations on its own. What you’d actually be doing is statistical modeling and Monte Carlo simulation — which is very real and powerful, but requires defining a model first. The AI helps you build and run those simulations, not conjure them from raw literature.

SETTING UP CLAUDE PRO + CLAUDE CODE FOR TRT RESEARCH — COST & SETUP GUIDE
COST
Quick heads up: There was a pricing controversy on April 22, 2026. Anthropic briefly moved Claude Code to the $100-$200/month Max plans only, which caused a firestorm online — but they reversed it within hours. Claude Code is still included in the $20/month Pro plan.
Claude Pro — $20/month — includes Claude Code, 5x usage vs free, priority access, and projects. Claude Max 5x — $100/month — only needed if you hit Pro limits regularly. Start with Pro at $20/month and upgrade only if you hit usage limits during heavy simulation work.

SETUP GUIDE: AVOIDING “GARBAGE IN, GARBAGE OUT”
The quality of your output depends entirely on the quality of your inputs. Here’s how to do it right.

STEP 1 — SET UP CLAUDE PRO & A PROJECT. Go to claude.ai and upgrade to Pro. Create a Project (sidebar > New Project) — this gives Claude persistent context across all conversations. In the Project instructions, paste a system prompt like this: “You are a research assistant specializing in endocrinology literature analysis. Always cite study limitations. Flag conflicts of interest, small sample sizes (under 50), study duration under 12 weeks, and industry funding. Never extrapolate beyond what data supports.” This primes every conversation with your quality standards automatically.

STEP 2 — BUILD A CLEAN LITERATURE DATABASE. This is the most critical phase. Bad data here poisons everything downstream. Where to get papers: PubMed Central (pubmed.ncbi.nlm.nih.gov) is free and peer-reviewed only. The Cochrane Library is the gold standard for systematic reviews. ClinicalTrials.gov has raw trial data with less spin than published papers. Avoid blog posts, supplement company sites, and non-peer-reviewed sources. When uploading each paper to Claude, use this prompt: “Analyze this study and extract: (1) Sample size and demographics, (2) Duration, (3) Protocol/dosing, (4) Primary outcomes measured, (5) Funding source, (6) Conflicts of interest, (7) Methodology grade (RCT/observational/case study), (8) Key findings, (9) Limitations acknowledged by authors, (10) Limitations NOT acknowledged by authors. Format as JSON.” This builds a clean, consistent database rather than vague summaries.

STEP 3 — INSTALL CLAUDE CODE FOR SIMULATIONS. Install Node.js from nodejs.org. Open your terminal and run: npm install -g @anthropic-ai/claude-code. Then run “claude” and log in with your Pro account. Ask Claude Code to build a Python script that reads your JSON literature database, construct statistical distributions from real study data (effect sizes, variance, etc.), run Monte Carlo simulations using those distributions, and flag which variables most robustly predict outcomes across simulations. Key GIGO safeguard: tell Claude Code to source every model assumption directly from your literature database, not from general knowledge. This keeps simulations grounded in actual data, not AI assumptions.

STEP 4 — QUALITY CONTROL LAYERS. Cross-reference everything: when Claude summarizes a paper, spot-check 2-3 claims against the actual PDF. Ask for uncertainty: prompt Claude to always include confidence levels (“How confident are you, and what would change your assessment?”). Run the same question multiple ways: slightly rephrase key questions and compare answers — inconsistencies reveal weak data or hallucination. Separate tiers of evidence: have Claude tag every finding as RCT, meta-analysis, observational, or case study, and weight simulations accordingly.

REALISTIC TIMELINE AND COST. Literature collection takes 1-2 weeks using PubMed (free) and Claude Pro ($20). Structured extraction takes 2-3 weeks using Claude Pro ($20). Simulation build takes 1-2 weeks using Claude Code which is included in Pro at no extra cost. Simulation runs and analysis are ongoing using Python (free) and Claude Pro ($20/month). Total cost is $20/month. The heavy computation runs locally in Python on your own computer for free — Claude Code just writes and iterates the code for you.

Nelson Vergel · Apr 23, 2026

Not a bad idea!

Guided_by_Voices · Apr 23, 2026

I'm in a hurry here but I'm skeptical. I have seen a lot of misleading or wrong information on health topics in general and TRT in particular from AI. Two days ago I looked into something and the AI response contradicted the first reference. Health and TRT in particular is subject to so much mainstream bias that I suspect it would take more human time to fix the AI wrongness than what it would take to generate by a human. We have seen the controversy over starting doses so I can only imagine the nightmare that TRT+ would be....I have learned that questions have to be asked in a way that requires a lot of prior knowledge so that it basically is confirming what you already know (or are close to knowing). Otherwise it will not volunteer obvious information that would have answered your question.

Phil Goodman · Apr 23, 2026

Guided_by_Voices said:
I'm in a hurry here but I'm skeptical. I have seen a lot of misleading or wrong information on health topics in general and TRT in particular from AI. Two days ago I looked into something and the AI response contradicted the first reference. Health and TRT in particular is subject to so much mainstream bias that I suspect it would take more human time to fix the AI wrongness than what it would take to generate by a human. We have seen the controversy over starting doses so I can only imagine the nightmare that TRT+ would be....I have learned that questions have to be asked in a way that requires a lot of prior knowledge so that it basically is confirming what you already know (or are close to knowing). Otherwise it will not volunteer obvious information that would have answered your question.

I agree to a large extent with your concerns and share them myself. I would note though, there is a pretty big difference (to put it lightly) between a general free chat platform and setting up an Agent to run a project. And the gap is growing with each iteration. For reference you can look into Project Glasswing and see that the newest version is so powerful that Anthropic is giving major companies first access so they can use it to secure their environment before releasing it to others, if it even ever does see wide scale availability. That is a somewhat different aspect though, as that all relates to cybersecurity which agents are better at than the complexities of the human body, but the point about improvements still remain.

On that note, there are stories like this… where an agent was used to help someone develop a cancer vaccine for their dog. Obviously there was a lot more that went into it, but it wouldn’t have been possible without the amount of lifting AI was able to take on during the process.

Man's dog was riddled with tumors and dying. He used ChatGPT to design a custom cancer vaccine, stunning researchers

With no biology degree and just $3,000, he designed a personalized mRNA vaccine in a world first. Moderna and Merck are spending billions to bring the same technology to human patients.

finance.yahoo.com

1. Synthesizing complex research

ChatGPT helped him:

Summarize scientific papers on mRNA cancer vaccines
Explain concepts like tumor mutations, neoantigens, and immune targeting
Translate dense biology into something actionable

Think of this as compressing weeks of literature review into hours.

2. Generating hypotheses and next steps

It suggested:

Exploring a personalized vaccine approach based on the dog’s tumor genetics
Using sequencing data to identify targets (neoantigens)
General workflows used in experimental oncology

Important distinction:
These are existing scientific approaches—ChatGPT didn’t invent a new therapy.

3. Helping structure a workflow

It assisted in mapping out steps like:

Sequence the tumor
Identify mutations
Predict which mutations could trigger an immune response
Design an mRNA construct targeting those mutations

That’s basically a simplified version of how experimental human cancer vaccines are designed.

4. Acting as an interactive sounding board

He could:

Ask follow-up questions continuously
Refine ideas quickly
Sanity-check whether something aligned with known science

This is where ChatGPT shines—it’s like having a 24/7 explainer + brainstorming partner.

What ChatGPT did NOT do

This is the part that often gets exaggerated:

It did not run lab experiments
It did not validate safety or efficacy
It did not manufacture the vaccine
It did not replace experts

The actual vaccine required:

Lab sequencing
Bioinformatics tools beyond ChatGPT
Collaboration with university researchers
Real-world testing and monitoring

Bottom line

ChatGPT’s role was closer to:

“accelerating understanding and guiding exploration”

—not:

“designing a working cancer cure on its own.”

Based on things like that I think we may be able to parse out some interesting things, though again I’m sure there would be some bumps along the way, and possibly a learning curve.

Guided_by_Voices · Apr 23, 2026

I'll write more tonight or tomorrow, but I'm curious if the the agent above identified the protocol used by the Quest Cancer Sanctuary for Dogs which has apparently had great success (even though it wasn't asked). Asking the agent how to add a low dose of XXXXXXXXXX to TRT target visceral fat as safely as possible would be an interesting test, especially since a lot of the data on Tren when approved for humans was in French.

Seagal · Apr 23, 2026

What question would you like to ask such a TRT AI Agent?

When it comes to TRT I'm always annoyed by the lack of more fundamental research (understanding of core principles, validated biochem modells etc) and the lack of study data that applies to us.

I find the search function on excelmale disappointing. An AI Agent could probably sort through excelmale and somehow rate the posts and ultimately answer questions to all topics discussed based on current exelmale wisdom.

Phil Goodman · Apr 23, 2026

Seagal said:
What question would you like to ask such a TRT AI Agent?

When it comes to TRT I'm always annoyed by the lack of more fundamental research (understanding of core principles, validated biochem modells etc) and the lack of study data that applies to us.

I find the search function on excelmale disappointing. An AI Agent could probably sort through excelmale and somehow rate the posts and ultimately answer questions to all topics discussed based on current exelmale wisdom.

Not sure yet, probably start with something basic like what an ideal protocol would be with regard to frequency of injection and dose. In the beginning I’ll just focus on determining which studies are most strongly structured and what are the most well-established concepts that people in the trt world believe vs which are true but not as strongly supported by the literature. And obviously which protocols(dose, frequency, delivery method, etc.) are most likely to optimize benefits and health outcomes while minimizing risks.

After that I might scrape social media and forums to compile a database of protocols and the outcomes/side effects they’re associated with. Obviously that’s less scientific due to basically zero data validation in place, but I’d be interested to see how the results from that compare to the first project. That would also provide more data points and also more data as people could discuss other aspects like HCG, SHBG levels, and other variables that could possibly result in finding more commonalities in the trt world. Not really sure yet, still thinking about interesting things to research and the best ways to go about doing it.

Seagal · Apr 24, 2026

Phil Goodman said:
Not sure yet, probably start with something basic like what an ideal protocol would be with regard to frequency of injection and dose. In the beginning I’ll just focus on determining which studies are most strongly structured and what are the most well-established concepts that people in the trt world believe vs which are true but not as strongly supported by the literature. And obviously which protocols(dose, frequency, delivery method, etc.) are most likely to optimize benefits and health outcomes while minimizing risks.

After that I might scrape social media and forums to compile a database of protocols and the outcomes/side effects they’re associated with. Obviously that’s less scientific due to basically zero data validation in place, but I’d be interested to see how the results from that compare to the first project. That would also provide more data points and also more data as people could discuss other aspects like HCG, SHBG levels, and other variables that could possibly result in finding more commonalities in the trt world. Not really sure yet, still thinking about interesting things to research and the best ways to go about doing it.

Just do it!

Phil Goodman · Apr 24, 2026

Seagal said:
Just do it!

Started this morning and have almost worked my way through PubMed. Then gonna move on to Google Scholar, Cochrane Library, Clinical Trials.gov, and any others that might have valuable data.

Guided_by_Voices · Apr 24, 2026

Ok, here's a less-hasty response, and kudos to you for taking the initiative.

Regarding possible topics:

- We frequently see people here who are asking if they should start TRT who clearly have not had a proper work-up to determine if their symptom and/or low T are the result of something else that can be fixed such as chronic infections, poor nutrition, low thyroid. etc.. Thyroid is top of mind for me since I listened to an interesting podcast with a thyroid specialist recently who said that in her experience, most low T is coming from untreated thyroid issues which mainstream thyroid approaches do not address. So.... an agent that could ingest symptoms (and ideally bloodwork, although it would have to be very comprehensive) and describe what an integrative/holistic work-up should entail seems like it would fill an important need.

- While tangential to TRT, another underserved topic IMO is Prostate cancer prevention. The topic of prevention is frequently conflated with early detection, however preventing cancer such that there is not anything to detect is woefully underserved IMO. There is emerging evidence that some BPH and cancers are due to infections and leaky gut for example, so a really thorough compilation of actions to take serves another major need. I realize there would be a lot of overlap with cancer prevention in general, but that is manageable and likely a good thing.

Regarding rules for the Agent, which IMO is more important than the topic, I think the rules have to be structured to avoid mainstream medical bias and a US-centric bias. So rules would include having a global focus, including compounds not necessarily approved in the US, rationalizing guidance against goals/outcomes, de-emphasizing proprietary approaches, prioritizing data from forward-thinking clinicians, prioritizing outcomes over most aspects of bloodwork, and factoring in other body systems that may be affected by goals (such as Nandrolone use for joints.)

The caution I see regarding ideal frequency and dose is that it is so individualized and subject to experimentation.

Phil Goodman · Apr 24, 2026

Guided_by_Voices said:
Ok, here's a less-hasty response, and kudos to you for taking the initiative.

Regarding possible topics:

- We frequently see people here who are asking if they should start TRT who clearly have not had a proper work-up to determine if their symptom and/or low T are the result of something else that can be fixed such as chronic infections, poor nutrition, low thyroid. etc.. Thyroid is top of mind for me since I listened to an interesting podcast with a thyroid specialist recently who said that in her experience, most low T is coming from untreated thyroid issues which mainstream thyroid approaches do not address. So.... an agent that could ingest symptoms (and ideally bloodwork, although it would have to be very comprehensive) and describe what an integrative/holistic work-up should entail seems like it would fill an important need.

- While tangential to TRT, another underserved topic IMO is Prostate cancer prevention. The topic of prevention is frequently conflated with early detection, however preventing cancer such that there is not anything to detect is woefully underserved IMO. There is emerging evidence that some BPH and cancers are due to infections and leaky gut for example, so a really thorough compilation of actions to take serves another major need. I realize there would be a lot of overlap with cancer prevention in general, but that is manageable and likely a good thing.

Regarding rules for the Agent, which IMO is more important than the topic, I think the rules have to be structured to avoid mainstream medical bias and a US-centric bias. So rules would include having a global focus, including compounds not necessarily approved in the US, rationalizing guidance against goals/outcomes, de-emphasizing proprietary approaches, prioritizing data from forward-thinking clinicians, prioritizing outcomes over most aspects of bloodwork, and factoring in other body systems that may be affected by goals (such as Nandrolone use for joints.)

The caution I see regarding ideal frequency and dose is that it is so individualized and subject to experimentation.

Once done with the trt project (assuming I'm not burned out lol) I might try to dive into the thyroid data.

And agreed with regard to the rules. I tried to make sure to weight the studies used for the initial database accurately. I can share prompts if interested, but this part is just a tiny piece of the puzzle. I actually have the master dataset with regard to studies done now (ended up being 113 different studies of varying confidence). And if anyone wants the .json file to save themselves the leg work on that front just let me know. The medical bias isn't really applicable at this point since it is all based on studies and not necessarily what doctors are doing. But that cuts both ways since we also aren't getting input from forward thinking clinicians to factor into the thought process. And pretty much all good studies done abroad are also included in literature that is searched for sources, so plenty of studies from all over the world are in the master dataset.

Gonna be a longer task than I'd originally planned. I asked for a general timeline along with the process just to have a better idea to plan ahead and got the response below. And this is assuming lots of full-time attention that I probably won't be able to dedicate, especially at this time of year. I am really interested in getting to the part where I scrape Reddit and other sources to see what signals start to emerge from first hand experience that isn't captured in studies though. On that note, if anyone has suggestions for where to pull from I'm open to them in order to get as much good data as possible.

WHAT SIMULATIONS ACTUALLY ADD THAT YOUR DATABASE CANNOT

Your database contains point estimates — study A found a 15% improvement in libido at this dose. But that single number hides enormous uncertainty. A Monte Carlo simulation takes all the reported effect sizes and variance across all your studies and asks a fundamentally different question: given the full range of what the literature shows, what outcomes are robust across thousands of hypothetical scenarios vs what outcomes only appear under ideal conditions? That distinction is genuinely new information that no single study or summary can tell you. It tells you which findings are fragile and which are solid regardless of individual variation.

Simulations also let you combine variables in ways no single study ever tested. No study has simultaneously examined injection frequency, estradiol management, baseline testosterone level, age, and body composition together. A simulation can model all of those interactions simultaneously based on what each individual study found about each variable separately.

WHAT YOU NEED TO GO BEYOND REGURGITATION

To make simulations genuinely generative rather than just reflective you need additional inputs beyond your current literature database. Here is what would meaningfully expand what the simulations can discover:

Individual patient level data — This is the gold standard. Instead of just knowing study averages you would have each individual participant's data points. Some researchers share this through platforms like the NCBI database or the Open Science Framework. Claude Code could search for any TRT trials that published their raw anonymized participant data and download it. This lets your simulation model individual variation rather than just group averages, which is enormously more powerful.

The social media database — This is actually the biggest opportunity and why your Phase 7 project is so valuable. Reddit and forum data gives you thousands of real world individual level reports with protocol details and outcomes. That is not clinical data but it represents a huge volume of natural variation that no clinical study captures. When you feed that into simulations alongside the literature you are now modeling something much closer to real world population behavior rather than controlled trial averages.

Pharmacokinetic modeling — Testosterone behaves in mathematically predictable ways in the body. There are published pharmacokinetic models for different delivery methods showing exactly how testosterone levels rise and fall with different doses and frequencies. Claude Code can implement these models directly, meaning the simulation can calculate predicted serum levels for any hypothetical protocol rather than just interpolating from study averages. This is genuinely new generative capability.

Bayesian updating — Rather than running simulations purely from your existing data, you can set up a Bayesian model that starts with your literature as the prior and then updates its predictions as you feed it new information — whether from new studies, social media data, or individual reports. This means the simulation actually gets smarter and more precise over time rather than just repeating itself.

WHAT THIS LOOKS LIKE PRACTICALLY

Phase 1 of simulation — Build from your current 113 study database. This establishes baseline distributions and identifies which findings are robust vs fragile. Acknowledge this phase is exploratory rather than generative.

Phase 2 of simulation — Feed in raw participant level data from any open access trials you can find. Claude Code searches for and downloads this automatically from Open Science Framework and similar repositories.

Phase 3 of simulation — Integrate your Reddit and forum database once Phase 7 is complete. Now your simulation is drawing on thousands of real world individual reports in addition to clinical literature.

Phase 4 of simulation — Implement pharmacokinetic models for different delivery methods. Now the simulation can model protocols that were never directly studied.

Phase 5 of simulation — Set up Bayesian updating so the model improves continuously as new data comes in.

THE HONEST BOTTOM LINE

Your current database alone would produce simulations that are useful for identifying robust vs fragile findings and modeling variable interactions, but you are right that they would not generate truly novel discoveries. The combination of your literature database, open access raw trial data, and the social media corpus from Phase 7 is where simulations start producing insights that genuinely could not be obtained any other way. That three-way combination is actually a novel research methodology that very few people are applying to this domain right now.

TIMELINE SUMMARY

Stage 1 full text gap filling — 1 to 2 weeks
Stage 2 deep dive analysis — 1 to 2 weeks running parallel to Stage 1
Stage 3 raw data collection — 1 week setup then ongoing
Stage 4 social media collection — setup this week then runs in background for 4 to 8 weeks
Stage 5 simulation build — 2 to 3 weeks after Stages 2 and 3 are solid
Stage 6 synthesis — 1 week after simulations are complete

Total realistic timeline from where you are now — 2 to 3 months for a genuinely comprehensive output.

Phil Goodman · Apr 25, 2026

Reddit scraper is up and running. Going to build some others and am open to suggestions in case anyone can think of other valuable resources to pull from. Also wanted to drop this here because @Nelson Vergel the last comment in the output should make you feel good about what you've built here.

Can You Scrape Other Sources?

Yes, but it varies by site:

ExcelMale.com — Very doable. It runs on a standard forum platform (XenForo) which has predictable URL structures. Ask me to build a scraper and I can write one using requests and BeautifulSoup that navigates forum threads and pages.

Other forums (Peak Testosterone, MESO-Rx, etc.) — Same approach, each needs a custom scraper but all are buildable with requests and BeautifulSoup.

Twitter/X — Very difficult now. API access is expensive ($100+/month) and scraping without credentials is actively blocked.

Facebook groups — Essentially impossible to scrape. Private groups require login and Facebook actively blocks automated access.

YouTube comments — Doable via the free YouTube Data API. Good source for TRT video comment sections.

Quora — Possible with requests but inconsistent. Lower value for your research than forums.

Best additional source after Reddit is ExcelMale — it's the highest quality TRT forum online with very detailed protocol discussions. Want me to build that scraper now?

Guided_by_Voices · Apr 25, 2026

If the allthingsmale site is still available via the wayback machine then I think it is worth considering. The "ultimate adnrenergic control of erections" thread is legendary in itself.

As an aside, I've gotten some of the best AI info (which I think is more accurately called "language enhanced search summaries") by asking it what a specific person thinks on a topic such as what Dr. Malcolm Hendrick thinks about a certain heart disease topic. That way the rules are somewhat built into the person being analyzed. For hormones Dr Mark Gordon and Dr Elizabeth Yurth might be a good candidates although that in itself creates a perspective bias.

Phil Goodman · Apr 25, 2026

Guided_by_Voices said:
If the allthingsmale site is still available via the wayback machine then I think it is worth considering. The "ultimate adnrenergic control of erections" thread is legendary in itself.

As an aside, I've gotten some of the best AI info (which I think is more accurately called "language enhanced search summaries") by asking it what a specific person thinks on a topic such as what Dr. Malcolm Hendrick thinks about a certain heart disease topic. That way the rules are somewhat built into the person being analyzed. For hormones Dr Mark Gordon and Dr Elizabeth Yurth might be a good candidates although that in itself creates a perspective bias.

Thanks for the forum suggestion, and honestly the way back machine may be easier to build a scraper for anyway.

Also, good thoughts on creating chats in the project to pull input from some of the leading doctors paving the way forward on this front. That can take care of that piece without introducing approaches from outdated doctors who still tell people trt will give you prostate cancer lol.

Phil Goodman · Apr 26, 2026

current status. Here are the files I have uploaded to the project so far:

TRT Research Repository — File Inventory

File	Size	Description
`trt_master_database.json`	69K	Core study registry — all 120+ studies with metadata, flags, quality ratings, paradoxes, evidence gaps
`trt_abstracts_master.json`	124K	Full abstract-level extractions across the corpus
`trt_safety_analysis_2026-04-25.json`	79K	Comprehensive safety domain analysis (CV, VTE, hematocrit, prostate, etc.)
`trt_responder_analysis.json`	81K	Responder heterogeneity — who responds, who doesn't, population gaps
`trt_delivery_method_analysis.json`	71K	Formulation comparisons: IM, SQ, transdermal, intranasal, oral, pellets
`trt_dosing_frequency_analysis.json`	71K	Dosing interval effects — PK, trough/peak ratios, frequency comparisons
`trt_e2_analysis.json`	57K	Estradiol domain — E2 targets, AI use, aromatization, E2 management
`hcg_domain_analysis.json`	53K	HCG co-administration — intratesticular T, fertility preservation, LH analog protocols
`phase3extractionA.json`	40K	Phase 3 deep extraction — batch A
`phase3extractionB.json`	49K	Phase 3 deep extraction — batch B
`phase3extractionC.json`	93K	Phase 3 deep extraction — batch C (largest batch)
`phase3mastersummary.json`	17K	Synthesis summary of all Phase 3 extractions — paradoxes, gaps, priority abstracts

Largest files: trt_abstracts_master.json (124K) and phase3extractionC.json (93K) — these carry the bulk of the raw extraction data.

Coverage domains represented: Cardiovascular safety, sexual function, body composition, bone density, metabolic outcomes, hematocrit, prostate, mood/cognition, delivery methods, dosing frequency, estradiol management, HCG protocols, special populations, and responder analysis.

If anyone can think of good questions to ask based on that information let me know. Shouldn't burn through much usage to have the agent analyze it for individual questions.

Currently building a datatbase using the information I've pulled from this forum, and at some point will compile one using the 2,600+ Reddit posts I've scraped. Also working on a pharmacokinetic script to include in the simulations then will move on to setting up the Monte Carlo simulations.

I just noticed this question:

Good afternoon madman! I am sitting wondering if I dare ask another question but I guess I will. I will return to the 100mg dose for six weeks but if you recall, I pushed my E2 to 50 while on 1x per week somewhere in the 8 week range. I'm not sure how your projection would have estimated this on 2x per week. Would I be better off or worse you think on 100mg 2x per or 3x per week? Just an opinion before I commit to the new protocol. Did you think I was a fast aromatizer at 100mg 1x per week? Thanks!

and I think these are exactly the types of things I hope to be able to weigh in on. Not to take anything away from Madman because he has been a Godsend for countless men coming to this forum asking questions (as evidenced by the thread the question was posed in), but I think having so much data condensed in one place coupled with a tool that can factor in so many variables and run countless simulations can hopefully be valuable (but may be a long way to go on that front).

Any input, questions, and/or suggestions are welcomed.