CancerGPT for few shot drug pair synergy prediction using large pretrained language models.

Large language models (LLMs) have been shown to have significant potential in few-shot learning across various fields, even with minimal training data. However, their ability to generalize to unseen tasks in more complex fields, such as biology and medicine has yet to be fully evaluated. LLMs can offer a promising alternative approach for biological inference, particularly in cases where structured data and sample size are limited, by extracting prior knowledge from text corpora. Here we report our proposed few-shot learning approach, which uses LLMs to predict the synergy of drug pairs in rare tissues that lack structured data and features. Our experiments, which involved seven rare tissues from different cancer types, demonstrate that the LLM-based prediction model achieves significant accuracy with very few or zero samples. Our proposed model, the CancerGPT (with ~ 124M parameters), is comparable to the larger fine-tuned GPT-3 model (with ~ 175B parameters). Our research contributes to tackling drug pair synergy prediction in rare tissues with limited data, and also advancing the use of LLMs for biological and medical inference tasks.

CancerGPT: Few-shot Drug Pair Synergy Prediction using Large Pre-trained Language Models.

Large pre-trained language models (LLMs) have been shown to have significant potential in few-shot learning across various fields, even with minimal training data. However, their ability to generalize to unseen tasks in more complex fields, such as biology, has yet to be fully evaluated. LLMs can offer a promising alternative approach for biological inference, particularly in cases where structured data and sample size are limited, by extracting prior knowledge from text corpora. Our proposed few-shot learning approach uses LLMs to predict the synergy of drug pairs in rare tissues that lack structured data and features. Our experiments, which involved seven rare tissues from different cancer types, demonstrated that the LLM-based prediction model achieved significant accuracy with very few or zero samples. Our proposed model, the CancerGPT (with ~ 124M parameters), was even comparable to the larger fine-tuned GPT-3 model (with ~ 175B parameters). Our research is the first to tackle drug pair synergy prediction in rare tissues with limited data. We are also the first to utilize an LLM-based prediction model for biological reaction prediction tasks.

Correction of Foot Deformities from Charcot Arthropathy with the Taylor Spatial Frame: A 7-14-year Follow-up.

Charcot arthropathy related foot and ankle deformities are a serious challenge. Surgical treatment of these deformities is now well established. Conventional surgical treatment includes extensive surgical exposure, excision of bone, acute correction and internal fixation, which is not always appropriate in presence of active ulceration, infection and poor bone quality. A minimally invasive approach to osteotomies and gradual correction of deformities using a circular frame are proving helpful in minimizing the complications. Taylor Spatial Frame (TSF) hexapod with its various modules is well suited for a range of foot and ankle deformities. We have advocated minimally invasive targeted hind and mid foot osteotomies and gradual correction with Taylor Spatial Frame (TSF) in 10 patients with recurrent ulceration and deformity. There are 2 female and 8 male patients in this cohort. Appropriate TSF module was chosen for each patient- a long bone module for ankle and hindfoot deformities (4 patients) and a forefoot 6x6 butt frame (6 patients) for foot deformities. An osteotomy through the midfoot was performed in all chronic stable foot deformity cases. In the ankle and hindfoot deformities, a combination of soft tissue distraction correction of equinus and acute correction of hindfoot deformity through a calcaneal osteotomy, were used. Our outcome measures are complete healing of the ulcers and resolution of infection, clinically plantigrade foot and ability to wear regular or diabetic footwear. Complications included eight episodes of pin infection that responded to oral antibiotics only and two pin breakages. We achieved ulcer and infection free plantigrade feet that fit in to regular or diabetic footwear in 9 out of 10 patients. 9 patients remain ulcer and infection free at a minimum of 7 years and maximum of 14 years follow up. Taylor Spatial Frame treatment provides an alternative to conventional surgery in high-risk complex Charcot neuroarthropathy foot and ankle deformities. How to cite this article: Lahoti O, Abhishetty N, Shetty S. Correction of Foot Deformities from Charcot Arthropathy with the Taylor Spatial Frame: A 7-14-year Follow-up. Strategies Trauma Limb Reconstr 2021;16(2):96-101.

Intentional deformation and closure of soft tissue defect in open tibial fractures with a taylor spatial frame--a simple technique.

OBJECTIVES: We describe a way of using Taylor spatial frames (TSFs) for acute deformation of Gustilo IIIB tibial fractures and infected tibial nonunions to close the soft tissue defects and to gradually restore anatomical alignment. We use the Direct Scheduler Utility module of the web-based software for TSFs to successfully restore the anatomical alignment. DESIGN: We report a case series of the above technique. SETTING: Care was carried out at a tertiary referral center for limb reconstruction. PATIENTS: Seven consecutive patients with significant soft tissue loss after tibial fractures were treated. INTERVENTION: Defects were closed with acute deformation of the fracture followed by gradual correction to anatomical alignment with a 2-ring TSF. MAIN OUTCOME MEASUREMENT: Successful closure of the soft tissue defect was our primary outcome measure. RESULTS: Soft tissue defects ranged from 3 to 10 cm. All healed without additional plastic surgery to cover the exposed tibia. Only 1 patient required an additional TSF prescription to achieve anatomical alignment. CONCLUSIONS: The use of the Direct Scheduler module of the web-based TSF software allows the complete correction of complex deformities without the need for obtaining complex mounting and frame parameters. We achieved successful closure of soft tissue defects and restored the anatomical tibial alignment in all our cases.