Skip to content
TheChatArchiveThe Chat Archive
Claudeclaude-instant-1· AnthropicHugging Face Arena Human Preference 55K

Hi! Could you please give some feedback on the following text? It is the discussion part…

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with Metallama-2-70b-chat’s answer.
  1. User

    Hi! Could you please give some feedback on the following text? It is the discussion part of a scientific article. Here we present a method that can measure the solubility of peptides in an high-throughput, inexpensive and label-free manner. It is high- throughput since there is only six steps and ev- ery step is done in a 96-well format. The six steps are: 1. distributing concentration series in a 96-well plate, 2. incubating the plate in a, for the studied system, suitable time and tempera- ture, 3. move the samples to 96-well filter plates, 4. pull sample through filter with centrifuga- tion (or vacuum), 5. move filtrate to a new 96- well plate prepared with OPA in each well, 6. measure fluorescence after 10-20 min. Step 4 and 5 can be combined by having OPA in the wells of the collection plate, but this gives less control over the volume of filtrate reacting with the OPA and reduces the possibility of quality control of the filtrates. The method is inexpen- sive since the only instrumentation needed is a centrifuge/vacuum manifold and a plate reader while the material needed for 96 data points is just 3 PEG-ylated 96-well plates, 1 filter plate, 480 μl OPA and of course buffer and protein. Which totals to ca 62 € for 96 samples, exclud- ing buffer and protein. It is also a very unbi- asad method, since there is no need for any la- bels added to the peptide or special components added to the buffer, the only requirement being that no primary amines exist in the system other than those on the peptide. Because of it’s simple design, this method is easily adapted for systems with slightly differ- ent requirements. If aggregation is impractically slow, one can add seeds, shake the plate during the incubation or simply work with higher initial concentrations. The filter plates are available in several different pore sizes and materials if a par- ticular system of interest produce smaller aggre- gates or have different adsorption properties. If the solubility for a certain system is lower than the limit of quantification reported here, there is two ways this method still can be applicable,

    1. further optimization of the OPA reaction such as reagent composition, volume and plate reader settings, 2. change the system, e.g. to a lower temperature or higher pH which might give a higher solubility, a few measurements at such different conditions should allow for extrapola- tion to the conditions of interest. The latter is a very common approach in the field of protein folding, where proteins are unfolded by heat or denaturant and their stability calculated by ex- trapolation
  2. Claude Assistant

    Here is some feedback on the discussion section:

    • The overall structure and flow of the discussion is good. It starts by summarizing the key advantages of the method (high-throughput, inexpensive, label-free), then provides more details to support these claims.

    • The level of detail provided about each step is appropriate. Explaining the practical aspects like the 96-well format and materials/costs needed helps validate the claims.

    • The discussion of how the method can be adapted is a strength. Acknowledging potential limitations and ways to address them improves the robustness. Suggesting alternative approaches like changing conditions shows flexibility.

    • The writing could be tightened in a few places. For example "ca 62 €" could just state the amount to improve clarity. And "because of it's simple design" could be shortened.

    • A few minor grammatical errors like "an high-throughput" and "unbiadad" could be corrected.

    Overall the discussion logically supports the method presented in the article and considers adaptations and limitations well. Tightening a few points of writing would help improve readability and clarity. But the structure and content are strong for this type of scientific discussion section.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.