Excel for antibody engineering

This Excel sheet has been developed by Ian Wilkinson as a free to use tool for anyone with an interest in antibodies. The latest version of the Excel sheet can be downloaded above. It is free to download with no license or restrictions of use. Below I provide a bit of background about myself followed by detailed instructions on how to use this tool. It has been designed to be as simple as possible to use. If you like the tool, have suggestions for improvements, have found an error or just want to keep up to date with the latest developments and launches please follow me on LinkedIn by using the following link: www.linkedin.com/in/ian-wilkinson-62772527    

I've been working with recombinant antibodies for 20 years starting with a PhD on antibody structure and function. I spent several years in large pharma R&D learning the ropes with all the cutting edge therapeutic antibody developments (mutli-specifics, ADCs, BBB delivery) before moving into the CRO world and helping to build a well respected service provider for antibody sequencing, expression and engineering. Following the acquisition of that business in 2020 I've moved on to several other ventures that are briefly described below. Although this Excel tool for antibodies isn't directly linked to any of these businesses it is only possible to make this freely available because of their support!

Throughout my journey as a student, pharma scientist, service-provider and now reagent supplier I have used Excel as a tool to do things that many people wouldn't think of as possible in a piece of software most people think of as for accountants. I find it fun to build little tools and get great satisfaction from the huge gains in efficiency I have seen from my teams that have used my Excel sheets over the years. By sharing this tool and any future versions I hope to be able to help a much wider group of people who don't have access to more sophisticated systems or the background knowledge to perform some of these tasks by hand themselves.

Gamma Proteins is my latest business, launched in early 2024. We supply high quality recombinant Fc receptors (specifically FcγR and FcRn currently) in bulk quantities and at a fraction of the cost of all existing recombinant protein suppliers. See our website for more details: www.gammaproteins.com

mAbsolve developed the STR mutations for Fc silencing. These mutations are widely regarded as the most silent yet reported and are available to license on reasonable terms for any therapeutic antibody developers not requiring Fc effector functions and wanting to avoid associated tox risks. www.mabsolve.com

I work as a freelance consultant under the trade name mAbvice. I can provide support on a range of different pre-clinical activities typically involved in therapeutic antibody development. For more details see my website: www.mabvice.com

As part of my work at mAbvice I provide an in silico humanization service. To provide more detail on the service as well as a lot of useful background information on the process and terminology of antibody humanization I created a separate website: www.antibodyhumanization.com

This Excel tool has been designed to be as simple as possible to input data and see the results. However, there are a lot of calculations going on in the background. These have been hidden away to create a clean user interface but there are some simple rules you should follow when using this Excel sheet to avoid corrupting the hidden formulae:

   1. Cells coloured light blue are data input cells. You can enter your data into these either as free text or using a drop down box. All other cells have been locked as they contain formulae that should not be edited.

   2. Please never cut and paste data within the spreadsheet or use 'drag and drop'. Both of these processes will drag formulae associated with the original cell to a new position. This series of Excel sheets has interlinked formulae across all of them so moving one cell can have an unexpected influence on the output on other sheets. If you input data in the wrong cell please just delete it and re-enter the data in the correct cell rather than dragging it over to the new cell.

   3. To avoid the issues associated with the above I highly recommend becoming familiar with the 'paste values' option. Within the paste drop down menu you will see a paste values option.  When transferring data (e.g. sequences) from another Excel sheet to this one using copy/paste I highly recommend using the paste values option as it will strip out any formulae associated with the original data and just paste over the text.

I have done my best to make the system as accurate as possible. It has been tested on >1000 different antibody sequences extracted from public databases. Much of the sequence analysis and manipulation relies on the fact that despite their immense variability, antibody sequences have many similarities, such as highly conserved Cys and Trp amino acids. Within my formulae these amino acids and other motifs are used as anchor points to identify CDRs, start/end of variable domain or constant domains etc. This works very well for >98% of antibodies across the species I have tested: human, primate, mouse, rat, rabbit, hamster, llama/alpaca and chicken. I believe it will work for anything except shark, which have a different CDR structure.  However, if your antibody sequence is highly unusual (e.g. lacks a conserved Cys) or has mutations that the system does not recognise then it may report errors or inaccurate information.

This is the main sheet to input your data. The sequence information entered here is then analysed in various ways on the other sheets. 5 types of data must be entered:

  1. Antibody name. This can be any string of text or numbers. This sheet accomodates up to 100 antibodies in total.

  2. Sequence type. This is a drop down menu with two options: Variable only; Full IgG. If your sequence(s) only contains the VH/VL domains then select variable only. If your sequence(s) is a full length heavy chain and light chain then select Full IgG. Important note - this sheet has not been designed to handle all manner of different antibody designs (fragments, fusions, bispecifics etc). If you have a non-standard antibody format please just input the VH/VL domain sequences to avoid confusing the system. IgGs with Fc domain mutations are ok.

  3. VH or HC sequence. Please input the variable heavy or full heavy chain sequence of your antibody. Please use single letter amino acid code. If the sequence contains a mammalian signal peptide this is OK.

  4. VL or LC sequence. Please input the variable light or full light chain sequence of your antibody. Please use single letter amino acid code. If the sequence contains a mammalian signal peptide this is OK.

  5. CDR scheme. Using the drop down menu select Kabat, Chothia or IMGT for the CDR numbering scheme. If you are unsure which to use I would suggest Kabat simply because this is the original and still most popular scheme despite the fact many updated versions now exist.

Once the above information has been input the sheet will perform the following automated analysis:

  1. Signal peptide identification in columns G and L respectively for VH and VL. If no signal peptide is detected the columns will report n/a (not applicable). The presence of a signal peptide is detected if a Methionine (M) is present at the start of the sequence. It will then try to detect the end of the signal peptide and start of the VH or VL domain. This works for most signal peptides but is not perfect due to the high degree of variability in signal peptide length, sequence and the starting amino acids of the VH/VL domains. If the sheet is not correctly identifying the signal peptide and start of variable domain please manually remove the signal peptide from your sequence(s). 

  2. VH and VL sequences are separated out from signal peptide and constant domains and reported in columns H and M respectively. In the following column a validity check is performed on the VH/VL sequence. This is to try and verify whether the sequences look like a typical variable domain sequence. Assuming the sequences passes all these checks the validity box will turn green and report the sequence as valid. The following checks are performed:

      a. Checks for presence of conserved Cysteine at position 20/21/22 prior to CDR1. If this is not present an error will be reported and the box will turn yellow: "Missing first Cys".

      b. It will check for the presence of any illegal characters within the sequence, such as non-amino acid letters or full stops. If a non-amino acid character is identified an error will be reported and the box will turn yellow: "Illegal char".

      c: Checks for the presence of a highly conserved WGxG motif within the VH and a FGxG motif within VL. If this is not present it will report an error.

      d: Checks for the correct ending of VH and VL domain based on the number of amino acids present after the WGxG/FGxG motifs at the end of CDR3. If the VH or VL has been truncated an error will be reported and the box will turn yellow: "Illegal VH/VL end".

      e: Checks the length of the VH/VL sequence. If this is too short (suggesting part of the sequence is missing) then an error will be reported and the box will turn yellow: "Seq too short".

  3. If the sequence contains a full heavy chain and light chain then the species and isotype of the antibody will be automatically identified. This is based on the presence of characteristic sequence motifs within the constant domains. If your antibody sequence has an unusal sequence at one or more of these sites then the system may not be able to identify the species/isotype or may misidentify it. This identification works well for human IgGs but is slightly less reliable for other species. It has been coded to try and identify the following IgG subtypes:

     - Human IgG1, 2, 3 and 4

     - Mouse IgG1, 2a, 2b, 2c and 3

     - Rat IgG1, 2a, 2b and 2c

     - Rabbit IgG

     - Hamster IgG

     - Dog IgG1 and 2

The aim of this sheet is to extract CDR sequences and then analyse those CDRs for common sequence liabilities. No data can be entered into this sheet, everything is taken from the Seq input sheet. The antibody name, CDR scheme, VH/VL sequence and CDR scheme are the same as on the Seq input sheet. To alter the CDR scheme go to the Seq input sheet and select an alternative CDR scheme using the drop down box.

Next to the validity box are columns for 'Glyco' and 'Cys'. These analyse the whole VH or VL sequence for the presence of an N-linked glycosylation motif (NXS or NXT) or a free Cysteine. If either are present in any position across the entire variable domain then the box will change to a red colour, indicating a high risk liability, if not present the box will remain green.

Each CDR is then analysed independently for common liabilities: deamidation, isomerization, oxidation and cleavage. These are determined as follows:

  1. Deamidation motifs are classified as high (H), medium (M) or low (L) risk and their presence is indicated by the colour red, orange or yellow respectively. High risk deamidation motif is NG, medium risk is NS, NN, NT or NH and low risk is NQ, NF, NW or NY.

  2. Isomerization motifs are classified as high (H), medium (M) or low (L) risk and their presence is indicated by the colour red, orange or yellow respectively. High risk isomerization motif is DG, medium risk is DS, DD, DN or DR and low risk is DY.

  3. Oxidation motifs are simply determined by the presence of a Methionine (M) in the CDR and are coloured red if present.

  4. Cleavage motifs are classified as high (H) or medium (M) risk and their presence is indicated by the colour red or orange respectively. High risk cleavage motif is DP and medium risk is TS.

For simplicity and to avoid too much data/confusion the framework regions (i.e. the sequences between the CDRs) have not been analysed for liabilities other than glycosylation and free Cys. 

It should be noted that these are simply predictions based on sequence motifs and do not mean that any of these sites will actually be modified. As can be seen on the 'INN antibodies' sheet, many clinical stage antibodies contain sequence liabilities within the CDRs so their presence is not always a problem but this should be explored during R&D. The ranking of severity or likelihood of a motif being modified is also debatable with people often ranking these differently. I have largely based my ranking from what I believe to be the most extensive study of deamidation/isomerization of clinical stage antibodies which can be found at the below link:

https://doi.org/10.1080/19420862.2018.1548233

The aim of this sheet is to analyse the constant domains of the heavy and light chains. No data can be entered into this sheet, everything is taken from the Seq input sheet. If only variable domain sequences have been entered into the Seq input sheet then the C domain analysis sheet will not perform any analysis. For non-human IgG sequences the analysis will be limited to identification of the species and isotype only. For human IgGs the following additional analysis will be performed:

  1. The allotype will be identified based on the amino acids present at allotypic sites. The amino acids at these sites will be reported in the 'Allotype residues' column and then the naming of the allotype will be provided in the 'Allotype name' column and is based on the IMGT naming scheme. For a nice table of allotypes I highly recommend users to look at Figure 1 in the following paper: 

https://doi.org/10.3389/fimmu.2020.00740

  2. If mutations in the constant domains are present (i.e. deviations from wild type  human IgG1/2/3/4 sequence) then these will be identified in the HC/LC mutations columns in the format XnnnY, where X is the wild type amino acid, nnn is the EU position number within the constant domain and Y is the substituted amino acid (e.g. L234A). This tool can identify any number of mutations within a sequence.

  3. For the heavy chain only, the tool will then try to indicate the predicted impact of any mutations in the sequence as follows:

      a. Effector function. If the mutation(s) are expected to reduce ADCP/ADCC/CDC then a red downwards arrow (↓) will appear in the column. If the mutation(s) are expected to increase ADCP/ADCC/CDC then a green upwards arrow (↑) will apear in the column. Note that this is not a prediction of the relative activity of an Fc domain. Both a wild type IgG1 and IgG4 will have no arrows for effector function despite the fact IgG4 has naturally lower effector function. The arrows simply indicate mutations that are likely to have an impact above and beyond what is expected for the wild type sequence.

      b. Half-life. If the mutation(s) are expected to increase plasma half-life then a green upwards arrow (↑) will appear in the column. If the mutations are expected to decrease half-life then a downwards red arrow will appear (↓). 

      c. Multispecific. Although this tool has not been designed to handle multispecific antibody sequences, if any mutations are present that are typically found in multispecifics then this will be indicated with a tick (P) in the column.

      d. Stabilization. If any mutations are present that stabilize the antibody (e.g. S228P is very common in IgG4s) then this will be indicated with a tick (P) in the column.

      e. C-term truncation. During production the C-terminal lysine of antibodies is commonly cleaved off the end of the heavy chain. This has led to some developers producing antibody sequences without this C-terminal lysine (K447), while others retain it within the sequence. If the lysine has been removed from the sequence this will be indicated with a tick (P) in the column. 

Important note - if your antibody sequence contains an insertion or deletion (e.g. G236 is occasionally deleted from human IgG1 for the purpose of Fc silencing) then this will cause havoc with the sequence alignment that is performed in the background and shift your sequence out of frame from the reference sequence. This will likely result in every amino acid after the insertion/deletion being identified as a mutation in error.

Although all the mutations present in a sequence should be identified the system may not correctly catagorise them all. I have tried to incorporate all the common mutations into this but some less common mutations may be identified but not reported as impacting effector function, half-life etc. I have incorporated almost all the mutations we identified in our large scale analysis of INN antibody sequences that was reported here:

https://doi.org/10.1080/19420862.2022.2123299

This sheet provides a simple tool to transfer the variable domains of any antibody onto an alternative isotype or species, such as taking an original human IgG1 and converting to human IgG4, or mouse IgG2a etc.

Data must first be entered into the 'Seq input' sheet. The name of the antibody then appears on the 'Chimerization tool' sheet. In the 'Desired heavy chain' column C you can select the type of antibody you wish to switch to using the drop-down menu. The new heavy chain sequence with your original VH domain and the new constant domains will then populate column F.

The correct light chain sequence will be auto-determined based on the original sequence (i.e. kappa or lambda) and the species of the new heavy chain that has been selected.

This tool allows for the introduction of mutations (substitutions only) into any human IgG heavy chain.

Sequence information should be entered into the 'Seq input' sheet as usual. The ten blue columns named Mut-1 to Mut-10 are then for the introduction of up to 10 mutations into your sequence. Within each blue box enter the mutation you want to incorporate in the format XnnnY, where X is the wild type amino acid, nnn is the EU position number within the constant domain and Y is the substituted amino acid (e.g. L234A). You will know whether the mutation has been successfully introduced in two ways:

  1. The number of mutations column will count the number of mutations in the new sequence.

  2. If you scroll to the right you will see an alignment of the original sequence with the mutated sequence along with the EU heavy chain numbering of CH1, hinge, CH2 and CH3 domains. Any differences between the two sequences will be indicated with a purple colouring below the mutation.

If your mutations have not been incorporated then it is likely that you have got the original amino acid and or position number incorrect. For instance, in the above pictured example the correct entries in Mut-1 and Mut-2 boxes are L234A and L235A. If these had been input as 234A, L236A, G234A or any other incorrect versions then the mutations would not be introduced.

Due to the complexity of this sheet the newly created mutated sequences are not on consecutive rows and so extracting multiple new sequences would be cumbersome. To aid with this I have created a 'Mutation output' sheet which is a simplified view of the outputs from the Mutation tool. From here you can quickly see and extract all the mutated sequences along with a summary of the number and type of mutations that have been incorporated.

A reasonable amount of the functionality within this spreadsheet was originally built for the analysis work that I performed as part of the review paper that Geoff Hale and I published: 'Systematic analysis of the varied designs of 819 therapeutic antibodies and Fc fusion proteins assigned international nonproprietary names' (linked below). I have then built the system out to be even more useful.

The 'INN antibodies' sheet contains 599 IgGs identified in that publication that have been used to develop the underlying formulae in this Excel tool. There are 19 antibodies (approx 3%) where my tool can't perform a complete analysis, usually because the sequence of the antibody is highly unusual. 

This sheet allows users to see the variable and constant domain analysis of a large number of clinically relevant IgGs. The only thing that can be changed on this sheet is the CDR scheme using the drop down menu.

https://doi.org/10.1080/19420862.2022.2123299

The reformatter has been developed to allow the quick and easy reformatting of up to 3 different VH/VL pairs into almost any antibody design, including the introduction of appropriate mutations to create multispecifics and/or modulate effector function and half-life. The tool has been designed to be as simple as possible to use with no real antibody engineering experience required but it is important to read the following instructions to ensure you use the tool correctly otherwise you could end up with the wrong output sequences.

To better understand the different format designs please download the spreadsheet which contains images for each format.

The first step in using the tool is selecting the antibody format you want. There are 65 different designs that have been built into the tool. Simple cartoon images of each format are shown on the 'Format images' sheet and more detailed depictions are shown at the bottom of the 'Reformatter - instructions' page. If this is your first time using this tool for a particular format I would strongly recommend looking at the more detailed images below to understand the designs and where your VH and VL sequences will be placed in the final sequences. 

To select your desired format use the drop down box in column B. Once selected the number of specificities for that format will be shown in column C (represented by the symbol of an arrow hitting a target) and the valency will be shown in column D (represented by a chain link symbol).

Depending on the format that is selected some of the VH/VL cells may be blacked out as these are not requred for that particular format. Input your sequences into the appropriate boxes. If designing a mulltispecific take note of the detailed images at the bottom of this page to understand which VH/VL pair will be inserted where in the final design. If using a format with a camelid VHH (nanobody) just insert the VHH sequence in the VH box.

Only use variable domain sequences not full length sequences. If you are unsure of this use the Seq input tool to identify the VH and VL sequences for your antibody. The 'Seq validity' column will perform a quick check on your sequence to ensure there are no errors (e.g. non-amino acid characters, gaps and other sequence anomalies). If the box is green please proceed to the next step, if the box is red please use the 'Seq validity' tool to check your sequence for errors.

At the moment this tool has only been designed to create antibody formats with human IgG1 constant domains. Please select human in the drop down box or Not applicable (which will appear for any formats not containing constant domains). 

If you were wishig to design a human IgG4 as an effector null option I would recommend using a human IgG1 with silencing mutations, which this tool can design. See the effector function instructions below for further details on the options for Fc silencing.

If you wanted to design mouse versions this is not currently possible but it may be included in future versions of this tool.

If the format does not contain a scFv this box will just have the 'Not applicable' option. Please select this. If the format contains any scFvs you will have the option to design the scFv as either VH first or VL first. This will apply to all scFvs within the format.

Orientation of the scFv can have an impact on expression and stability of the scFv. It is not possible to predict which is better but as a rule I would recommend VL-VH orientation as this provides a shorter distance for the linker to span an so creates a less strained scFv.

If the format does not contain a scFv this box will just have the 'Not applicable' option. Please select this. If the format contains any scFvs you will have the option to select the linker length. All options incorporate a Glycine-Serine linker and are either 15, 20 or 25 amino acids long. If unsure which to pick I would recommend a 15 amino acid linker, i.e. 3x Gly4Ser.

If the format contains an Fc domain this box will just have the 'Not applicable' option. Please select this. If the format does not include an Fc domain then you may wish to incorporate a tag to aid purification and/or detection. You can select to include no tag ('None') or a variety of different options that have been built into this tool. Specifically the sequences of these are:

 - 6xHis has the sequence HHHHHH

 - 10xHis has the sequence HHHHHHHHHH

 - FLAG has the sequence DYKDDDDK

 - FLAG-His has the sequence DYKDDDDKGGGGHHHHHH

 - Strep-II has the sequence SAWSHPQFEK

If the format contains a standard homodimeric Fc then this box will just have the 'Not applicable' option. Please select this. If the format contains a heterodimeric Fc (i.e. an asymmetric format) then you will have the choice of 4 different sets of mutations to favour correct heavy chain pairing.

 - KiH. This is the original knobs into holes mutations. Specifically T366W on one side and T366S/L368A/Y407V on the other. For further details see the original publication, https://doi.org/10.1093/protein/9.7.617

 - KiH + Cys. This is the updated version of knobs-into-holes with a disulphide. Specifically S354C/T366W on one side and Y349C/T366C/L368A/Y407V on the order. For further details see the original publication, https://doi.org/10.1038/nbt0798-677

 - Duobody. Note that this design requires separate production of the two antibodies followed by a redox reaction to form the heterodimer. One heavy chain contains the K409R mutation and the other contains F405L. See the original publication for further details, https://doi.org/10.1073/pnas.1220145110

 - ES steering. This variant uses electrostatic steering mutations to drive correct heterodimer formation using mutations Q295E/L368D/K370S/N384D/Q418E/N421D on one side and E357Q/S364K on the other. See the original publication for further details, https://doi.org/10.1016/j.ymeth.2018.10.006

If the format contains a light chain then this box will just have the 'Not applicable' option. Please select this. If the format contains multiple light chains that need to be paired correct then you will have the choice of 3 different sets of mutations to favour correct light chain pairing.

 - Crossmab. This approach consists of swapping the CH1 and CL domains onto the opposing arm with careful consideration of how to connect the domains at the variable-constant domain hinge region. This tool allows the use of Crossmab for both kappa and lambda light chains. For further details please see the original publication, https://doi.org/10.1073/pnas.1019002108

 - Golay. This design uses CH1 and CL mutations to drive correct pairing. Specifically these are L145Q/S183V in the CH1 and V133T/S176V in the CL. It should be noted that these mutations only work for kappa light chains. For more details see the original publication, https://doi.org/10.4049/jimmunol.1501592

 - Duetmab. This design switches the position of the HC-LC disulphide to drive correct pairing. Specicially the mutations are F126C/C220S in the CH1 and S121C/C214S in the CL. These mutations work for both kappa and lambda light chains. For more details see the original publication, https://doi.org/10.1080/19420862.2015.1007816

If the format contains an Fc domain you will have to make a choice of whether to include mutations to modulate Fc effector function (ADCC, ADCP and/or CDC). If you wish to have a standard human IgG1 please select 'Wild type'. 

For Fc silencing (i.e. reduced binding to Fc gamma receptors and C1q) you have 4 options:

 - STR, containing mutations L234S/L235T/G236R. See the original publication for further details, https://doi.org/10.1371/journal.pone.0260954

 - LALA, containing mutations L234A/L235A.

 - LALAPG, containing mutations L234A/L235A/P329G. See the original publication for further details, https://doi.org/10.1093/protein/gzw040

 - N297A. This contains a single point mutation that removes the N-linked lycosylation motif and so results in an aglycosylated antibody.

For a general review of all the different silencing mutations that have been reported I recommend the following paper two publications:

https://doi.org/10.1080/19420862.2024.2402701

https://doi.org/10.1111/imr.13379

For enhancing effector function you have 5 options:

 - Hexabody. This contains the mutation E430G which promotes IgG hexamerization when engaged with a target on the cell surface and thus enhances complement activation (CDC) in particular.

 - GASDALIE. This contains the mutations G236A/S239D/A330L/I332E. These mutations enhance both ADCC and ADCP.

 - EFT+AE. This contains the mutations G236A/S267E/H268F/S324T/I332E. These mutations enhance CDC specifially. See the original publication for more details, https://doi.org/10.4161/mabs.2.2.11158

 - DE. This contains the mutations S239D/I332E. This greatly enhances ADCC and to a lesser extent ADCP. See the original publication for more details, https://doi.org/10.4161/mabs.2.2.11158

 - FT+DE. This contains the mutations S239D/H268F/S324T/I332E. These mutations enhance CDC, ADCC and ADCP to varying degrees. See the original publication for more details, https://doi.org/10.4161/mabs.2.2.11158

For a general review on the different Fc enhancing mutations I would recommend the following paper: https://doi.org/10.1080/19420862.2024.2406539

If the format contains an Fc domain you will have to make a choice of whether to include mutations to modify FcRn binding and thus half-life (PK). If you wish to have a standard human IgG1 please select 'Wild type'. 

For half-life extension please select one of the following:

 - YTE. This incorporates the mutations M252Y/S254T/T256E. See the original publication for further details, https://doi.org/10.4049/jimmunol.169.9.5171

 - LS. This incorporates the mutations M428L/N434S. See the original publication for further details, https://doi.org/10.1038/nbt.1601

 - QL. This incorporates the mutations T250Q/M428L. See the original publication for further details, https://doi.org/10.1074/jbc.c300470200

 - DHS. This incorporates the mutations L309D/Q311H/N434S. See the original publication for further details, https://doi.org/10.1038/s41467-019-13108-2

To eliminate FcRn binding and thus reduce half-life please select the following:

 - AAA. This incorporates the mutations I235A/H310A/H435A. See the original publication for further details, https://doi.org/10.1002/(SICI)1521-4141(199909)29:09%3C2819::AID-IMMU2819%3E3.0.CO;2-6

Depending on the format that is selected some of the sequence output boxes will be blacked out. The rest will contain full heavy and light chain sequences that incorporate your particular variable domains and any other mutations and other parameters that you have selected in the drop down boxes. Please ensure all drop down boxes have been selected (even if just to select Not applicable). The full molecular weight and extinction coefficient are also calculated for each molecule.

Although I have tested this tool extensively and found it to be highly accurate for most antibody sequences, no tool like this is perfect. I accept no responsibility for any errors in the system and inaccurate interpretation or mis-characterisation of your antibody sequences. This tool is designed to enable non-specialists to quickly analyse mutliple antibody sequences in parallel but I encourage all users to carefully check the validity of the data provided by my tool by speaking with experts and/or comparing the outputs with other tools.

I have made this tool freely available so that experts and non-experts alike can analyse antibody sequences quickly and easily without the need for costly access to private offline or online software. I am happy for users to download the software and share with colleagues within and outside of your organisation as you see fit. All I ask in return is for people that like the system to connect with me on LinkedIn (https://www.linkedin.com/in/ian-wilkinson-62772527/) and share any comments they have (positive or negative) so that I can incorporate feedback into future versions.