Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
med-mastodon.com is one of the many independent Mastodon servers you can use to participate in the fediverse.
Medical community on Mastodon

Administered by:

Server stats:

423
active users

med-mastodon.com: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

#oncology

7 posts6 participants1 post today
Public
Hermann Mucke

#LLMs could guide precision #oncology biorxiv.org/node/4448173.abstr

bioRxiv · SensitiveCancerGPT: Leveraging Generative Large Language Model on Structured Omics Data to Optimize Drug Sensitivity PredictionObjective The fast accumulation of vast pharmacogenomics data of cancer cell lines provide unprecedented opportunities for drug sensitivity prediction (DSP), a crucial prerequisite for the advancement of precision oncology. Recently, Generative Large Language Models (LLM) have demonstrated performance and generalization prowess across diverse tasks in the field of natural language processing (NLP). However, the structured format of the pharmacogenomics data poses challenge for the utility of LLM in DSP. Therefore, the objective of this study is multi-fold: to adapt prompt engineering for structured pharmacogenomics data toward optimizing LLM’s DSP performance, to evaluate LLM’s generalization in real-world DSP scenarios, and to compare LLM’s DSP performance against that of state-of-the-science baselines. Methods We systematically investigated the capability of the Generative Pre-trained Transformer (GPT) as a DSP model on four publicly available benchmark pharmacogenomics datasets, which are stratified by five cancer tissue types of cell lines and encompass both oncology and non-oncology drugs. Essentially, the predictive landscape of GPT is assessed for effectiveness on the DSP task via four learning paradigms: zero-shot learning, few-shot learning, fine-tuning and clustering pretrained embeddings. To facilitate GPT in seamlessly processing the structured pharmacogenomics data, domain-specific novel prompt engineering is employed by implementing three prompt templates (i.e., Instruction, Instruction-Prefix, Cloze) and integrating pharmacogenomics-related features into the prompt. We validated GPT’s performance in diverse real-world DSP scenarios: cross-tissue generalization, blind tests, and analyses of drug-pathway associations and top sensitive/resistant cell lines. Furthermore, we conducted a comparative evaluation of GPT against multiple Transformer-based pretrained models and existing DSP baselines. Results Extensive experiments on the pharmacogenomics datasets across the five tissue cohorts demonstrate that fine-tuning GPT yields the best DSP performance (28% F1 increase, p-value= 0.0003) followed by clustering pretrained GPT embeddings (26% F1 increase, p-value= 0.0005), outperforming GPT in-context learning (i.e., few-shot). However, GPT in the zero-shot setting had a big F1 gap, resulting in the worst performance. Within the scope of prompt engineering, performance enhancement was achieved by directly instructing GPT about the DSP task and resorting to a concise context format (i.e., instruction-prefix), leading to F1 performance gain of 22% (p-value=0.02); while incorporation of drug-cell line prompt context derived from genomics and/or molecular features further boosted F1 score by 2%. Compared to state-of-the-science DSP baselines, GPT significantly asserted superior mean F1 performance (16% gain, p-value<0.05) on the GDSC dataset. In the crosstissue analysis, GPT showcased comparable generalizability to the within-tissue performances on the GDSC and PRISM datasets, while statistically significant F1 performance improvements on the CCLE (8%, p-value=0.001) and DrugComb (19%, p-value=0.009) datasets. Evaluation on the challenging blind tests suggests GPT’s competitiveness on the CCLE and DrugComb datasets compared to random splitting. Furthermore, analyses of the drug-pathway associations and log probabilities provided valuable insights that align with previous DSP findings. Conclusion The diverse experiment setups and in-depth analysis underscore the importance of generative LLM, such as GPT, as a viable in silico approach to guide precision oncology. Availability <https://github.com/bioIKEA/SensitiveCancerGPT> ### Competing Interest Statement The authors have declared no competing interest.
Public
WYM Mod

A pivotal change at IBA—Catherine Vandenborre steps in as CFO! Announced on April 7, 2025, in Louvain-la-Neuve, her appointment underscores IBA's commitment to advancing its financial strategy and leadership as it strengthens its oncology solutions and scientific services. Vandenborre's extensive expertise promises to guide the company in achieving its long-term objectives. Read more about this strategic move [here](finance.yahoo.com/news/iba-lon). #IBACFO #Leadership #Oncology #Finance #BusinessNews

Yahoo Finance · IBA long-term strategy presented at 2025 Capital Markets DayBy IBA SA
Public
franebleu

For the Mosuo, in the vastness of Nepal, the absence of the norms of sedentarized marriage and male domination may have contributed to society's evolution towards peacefulness.

Male domination + individualism may be significant factors favoring violence and preventing societies evolution towards peace and maturity.

researchgate.net/publication/3

Picture of a woman holding a cardboard in a general strike. "I am no longer accepting the things I cannot change. I am changing the things I cannot accept."
#Patriarchy#MaleDomination#RevolutionWontBeMasculinized
Public
franebleu

Memes associated to the judgement element "Does it save lives ?" :

Useful meme :
Put a mask when there's a massive epidemic. (saves lives)

Neutral meme :
Use a tool to eat. (no big deal if you have clean hands)

Harmful meme :
It is impossible to put Elon Musk in prison immediately for the rest of his life, and it is impossible to seize his whole wealth and redistribute it to each US citizen through public services. (costs lives)

#Oncology
#NeutralizeAndRecycleBillionaires
#HealEarth
#Health

Meme Picture of Earth with rainbow colors and a massive purplish cancerous tumor in the background. Writings in pink and celeste. " - But Daddy... Then Earth and Moon are like some giant intelligent organism ? " " - Huh... Yeah. " " - Then... what are billionaires ?" " - Hmmm... Tumors ?" #HEALEARTH
Public *
halama_immuno

Very nice presentation on ligandomics-based treatment selection for personalized vaccination approaches. Prof Juliane Waltz summarised the state-of-the-art...with adjuvant application for #peptide or #mrna vaccines in #Oncology being the current field with promising results. #Immunotherapy combinations bring up the questions: when? What target? And so on... Plenty of aspects to explore! Immunotherapy of cancer conference #ITOC11 #immunology #science #health

Prof Juliane Waltz talking at the ITOC vonference
Overview of the steps in vaccination
Welcome banner #ITOC11
Replied in thread
Public *
franebleu

@JoeGrowling
Cool, but you don't remove parasites by taxing them, it doesn't work like that.

Ask any cancer survivor, 100% of them didn't tax their tumors. The big ones have to be removed, their remains have to be burnt alive, all their satellites have to be cleansed with chemical extermination.

This may leave big holes and cost you some hair, but everything gets replaced by sane structures in the end and it saves your life.

Taxes only prevent their return.

#Parasites
#Oncology
#Health

Public *
ScienceOpen

'Utility of Assessing Early Tumor Shrinkage as an Efficacy Predictor in Patients with Non-Surgically Indicated or Recurrent Esophageal Cancer Treated with Nivolumab plus Ipilimumab' - an #Oncology article by Karger Publishers on #ScienceOpen:
scienceopen.com/document?vid=8

ScienceOpenUtility of Assessing Early Tumor Shrinkage as an Efficacy Predictor in Patients with Non-Surgically Indicated or Recurrent Esophageal Cancer Treated with Nivolumab plus Ipilimumab<p xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" dir="auto" id="d3734252e235"> <b> <i>Introduction:</i> </b> Nivolumab plus ipilimumab combination therapy has been administered as a first-line treatment in Japan since 2022 for patients with unresectable progressive or recurrent esophageal cancer. The efficacy and safety of this immune checkpoint inhibitor (ICI) doublet therapy are now being evaluated, and it is necessary to identify populations that benefit from this treatment at an early phase after initiation. For patients not showing early benefit, changing as soon as possible to other therapeutic strategies could improve their survival outcomes. Therefore, we attempted to identify decision-making factors such as early tumor shrinkage (ETS) based on treatment experience with ICI doublet therapy. <b> <i>Methods:</i> </b> The study included 19 patients who received nivolumab plus ipilimumab for non-surgically indicated or recurrent esophageal cancer between July 2022 and November 2023. Tumors were assessed approximately every 2 months after treatment initiation. The effects of ETS, depth of response (DpR), and clinicopathologic features, including immune-related adverse events (irAEs), on progression-free and overall survival were evaluated using Kaplan-Meier plots and Cox proportional hazard models. <b> <i>Results:</i> </b> The mean duration of ICI doublet administration was 5.89 months (range, 1–16 months). At first evaluation, patients who exhibited no tumor progression >20% indicated possible response to ICI doublet therapy, and patients whose tumors shrank even minimally exhibited favorable progression-free survival. Higher DpR at any cut-off line exhibited better progression-free survival than those with lower DpR. Fifteen patients experienced irAEs, with 13 of these patients experiencing irAEs within 3 months of treatment initiation. irAEs were associated with the efficacy of ICI doublet therapy, but efficacy could not be predicted based on early irAE experience. <b> <i>Conclusion:</i> </b> ETS-high, DpR-high, and irAEs might be associated with favorable responses to nivolumab plus ipilimumab. As a predictor of efficacy at an early phase, ETS >0% could be a deciding factor for continuing ICI doublet therapy. </p><p xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" dir="auto" id="d3734252e264">We studied the usefulness of evaluating early tumor shrinkage (ETS) as an efficacy predictor in patients with inoperable or recurrent esophageal cancer treated with nivolumab plus ipilimumab therapy. The immune checkpoint inhibitor (ICI) doublet therapy has been administered since 2022, and there are still not many patients who received this treatment all over the world. It is necessary to find well-responders to this treatment as soon as possible to improve outcomes and consider therapeutic strategies. We analyzed tumor shrinkage and clinicopathologic features in patients with inoperable or recurrent esophageal cancer treated with the treatment. ETS was defined as the percent decrease in the sum of the largest diameter of the target lesions at first evaluation after treatment initiation compared to that at baseline. Regarding ETS, those whose tumors shrank even minimally had favorable progression-free survival. Immune-related adverse events were also associated with the sensitivity and efficacy of ICI doublet therapy, but to find the efficacy at an early phase, ETS could be a decision-making factor for predicting the efficacy and continuation of ICI doublet therapy. </p>
Replied in thread
Public *
franebleu

@megatronicthronbanks @VedaDalsette @WatchForBones @idoubtit @megatronicthronbanks
For the Mosuo, in Nepal's vastness, the absence of the norms of sedentarized marriage and male domination may have contributed to society's evolution towards peacefulness.

Male domination + individualism may be significant factors favoring violence and preventing societies evolution towards peace and maturity.

researchgate.net/publication/3

Woman holding a cardboard in a general strike. "I am no longer accepting the things I cannot change. I am changing the things I cannot accept"
#Patriarchy#MaleDomination#RevolutionWontBeMasculinized
ExploreLive feeds
About