Machine learning optimization of candidate antibody yields highly diverse sub-nanomolar affinity antibody libraries.

Therapeutic antibodies are an important and rapidly growing drug modality. However, the design and discovery of early-stage antibody therapeutics remain a time and cost-intensive endeavor. Here we present an end-to-end Bayesian, language model-based method for designing large and diverse libraries of high-affinity single-chain variable fragments (scFvs) that are then empirically measured. In a head-to-head comparison with a directed evolution approach, we show that the best scFv generated from our method represents a 28.7-fold improvement in binding over the best scFv from the directed evolution. Additionally, 99% of designed scFvs in our most successful library are improvements over the initial candidate scFv. By comparing a library's predicted success to actual measurements, we demonstrate our method's ability to explore tradeoffs between library success and diversity. Results of our work highlight the significant impact machine learning models can have on scFv development. We expect our method to be broadly applicable and provide value to other protein engineering tasks.

Mercury concentrations in water and hybrid striped bass (Morone saxatilis × M. chrysops) muscle tissue samples collected from the Ohio River, USA.

We report on long-term aqueous mercury (Hg) measurements collected at fixed locations along the Ohio River, offer insights into patterns of water and fish tissue Hg levels, and calculate site-specific bioaccumulation factors (BAFs) along an extensive longitudinal basis. We examined the relationship between total recoverable Hg concentrations in water and fish samples collected from 12 locations on the mainstem Ohio River. Water samples were collected on a bimonthly basis from each location over a 6-year period preceding the collection of fish tissue samples. This abundance of data enabled us to calculate the long-term average aqueous Hg concentrations and approximate the lifetime aqueous Hg exposure experienced by fish, enabling the calculation of appropriate BAFs. Hybrid striped bass (HSB; Morone saxatilis × M. chrysops) were collected from the Ohio River, composited (three fish), and analyzed for Hg in muscle tissue from each location. Concentrations ranged from 0.2 to 0.4 mg/kg and 41.7% of all samples collected were higher than the US Environmental Protection Agency regulatory threshold of 0.3 mg Hg/kg wet weight. Hg levels generally increased with fish weight, length, and age. However, Hg concentration in the water was the strongest predictor of tissue concentrations. We found that both water and tissue concentrations increased with drainage area, albeit at different rates. This discrepancy in spatial patterns revealed that the bioaccumulation rate of methylmercury might not be consistent throughout the Ohio River mainstem. BAFs calculated at each location supported this finding, as values decreased with increasing drainage area. Our study serves to fill critical, previously identified data gaps and provides decision-makers with the information necessary to develop more appropriate BAF development and risk-management strategies.

Laser imaging and computer-aided design and computer-aided manufacture in prosthetics and orthotics.

Although Hanger Orthopedic Group, Inc., has been developing clinical protocols for its INSIGNIA scanner for more than 2 years, there are many applications that are currently in development and will be released over the next 2-year period after this publication. It is the goal of Hanger Orthopedic Group, Inc., to replace all plaster casting procedures with the laser scan and move toward a paperless environment where all images and documents are passed through its virtual network. INSIGNIA currently has five major production centers throughout the United States, which support more than 600 INSIGNIA certified clinicians. These clinicians staff more than 600 clinics in North America, all under the Hanger company name. The central fabrication service and the central design center processes hundreds of shapes per day (Fig. 27). So that any clinician in the field can use the expertise of the central designers and central fabricators to help with overflow or problems they might be having, the network that exists within Hanger is tied together and enhanced by INSIGNIA. Through virtual modification and centralization of these services, each patient receives the virtual collaboration of several clinicians with a total of years of experience. INSIGNIA has enhanced the patient experience. The enhancement is not only in removing the plaster from the process, but also in exposing each patient to the team of prosthetic experts working collaboratively behind the scenes. The rehabilitation industry continues to be bombarded with compliance paperwork and justifications. The INSIGNIA scan and resulting measurement reports give inherent strength to justifications based on volume change, surgical revisions, or tissue change. The files are kept in a data warehouse where they are vaulted and preserved presumably forever. Also, any of the shape graphics or measurement instruments can be printed into a discrete report that can become part of the patient's permanent record. Many physicians receive update letters from their orthotic and prosthetic clinician with a status update before and after treatment of their patient. This update includes a descriptive narrative, a printout of the pertinent metrics, a printout of the scan graphic, and often a digital image of the patient wearing the device (Fig. 28). The network is HIPAA compliant, and all private health information is held in tight security. If a practitioner does not have a HIPAA agreement in place with Hanger Orthopedic Group, Inc., and would like one, or ifa practitioner would like to have an INSIGNIA representative call or visit with more information, the practitioner is encouraged to call 1-800-4-HANGER and request an INSIGNIA in-service or visit INSIGNIA on the web at www.hanger.com.