Why is KNIME so popular in the pharmaceutical industry?

I read up on KNIME and found out that it is especially popular in the pharmaceutical industry.

Since I have been using Alteryx for a long time, I wonder what KNIME can do that Alteryx can’t really do for this industry?

Thank you!

If I had to hazard a guess it would for the same reasons that I prefer KNIME. KNIME is open source and can really be built upon to do just about any custom work you would like. Alteryx is more of a closed platform with a limited set of official tools. If node functionality doesn’t exist in KNIME, you can always build one yourself or expand on another users creation. It also integrates well with other existing programming language solutions better and can handle more robust image analysis. I also think it is more intuitive when you get to looping, modeling or ML.

I still have an active Alteryx license on my computers because you can’t get out of their inflexible multi-year contracts, but haven’t even opened it in over a year. The platform feels limited and slow for me now. You can also get KNIME server for the cost of a couple of desktop Alteryx licenses. That equates to a lot of organizational savings on the desktop side and servers offer way more capabilities and processing power for resource intensive tasks.

12 Likes

If you go to File->Install KNIME Extensions you will see two Community Extensions; one for Bioinformatics and Next Gen Sequencing, and another for Cheminformatics. These provide specific functionality and interfaces to systems within the pharmaceutical industry and have been a part of KNIME for as long as I can remember.

10 Likes

Hi @Robinvm96 and welcome to the forum! :slight_smile:

Just some additional points to what the previous posters already mentioned:

  • We are quite strong for the cheminformatics functionality, integrating a lot of functionality with the RDKit and Vernalis extension (which are open source)
  • besides that we we have a lot of connectors and nodes related to 3rd party tools like Schrödinger, CCG, ChemAxon, etc.
  • as KNIME AP is open source and always will be, you will not face the risk of a hard stop. In case you decide to discontinue our commercial product, you can always still run your workflows on the AP
9 Likes

Thanks for your answer! You have mentioned many points that generally speak for KNIME. In fact, I would like to understand the differences of KNIME and Alteryx as much as possible and also be able to compare it. Are there any other general points in favor of KNIME in general?

One point in particular stands out to me:
Is image analysis particularly widely used in the pharmaceutical industry?
I know Alteryx offers computer vision. It is included in the Intelligence Suite, which again costs money.

Hello @Robinvm96,

differences between Alteryx and KNIME and in general questions regarding moving from Alteryx to KNIME was already discussed many times (I’m not saying you shouldn’t discuss or ask related questions for specific area just pointing out that there already might be answers lying around). Here I share two forum topics you might find useful:

Additionally you can use forum search on your own. Simply type Alteryx and you’ll get many results :wink:

Welcome to Community!

Br,
Ivan

7 Likes

I would offer you my own take on the the matter. The three main points are

  • KNIME is a platform. Sounds simple but it means it will happily interact with all sorts of data sources and systems like Java, R, Python etc. as well as SQL-databases and Big Data
  • It is highly scalable: it can be used by beginners as well as experts and they all can interact with the business side (all the hippos coming together). It is low-code
  • it has a great community and support network. And the initial price tag is just great :slight_smile:

Then you could check out Reviews about KNIME and KNIME-Sever at Gartner:


Then I like to reference these examples from leading German companies about their use of KNIME and the building of a data science community:

Expanding the Field: The Data Solutions Space for Self-Service Business Users (Continental)

Sparking Data Literacy with KNIME and Making Better Decisions (Continental)

Five Takeaways from the First KNIME Meetup@Siemens

Driving a Citizen Data Scientist Approach (Siemens)

Manufacturers Bosch, ZF, and Fraunhofer Discuss Key Considerations for Building Data Science Teams

8 Likes

I will add as well that in my experience KNIME is far more responsive to user feedback, and much quicker to incorporate them into updates. Alteryx has more resources at their disposal due to the higher price tag, but didn’t prioritize enhancements based on user feedback enough in my book. They fixed functional error problems, but seemed to develop their own version of “platform enhancements” internally while simple (but necessary) user requests going back a decade would just sit indefinitely.

It always comes down to culture in the end. KNIME has a better culture of utilizing user feedback to drive platform changes.

Those issues get magnified by the basic fact that…

Alteryx is a rigid “closed” system. If you hit a limitation wall, then you are out of luck unless they ever get around to changing things.

KNIME is an “open” system. If you hit a limitation wall, then there are typically several paths forward and a great community to help guide you. Just about every wall is scalable in one way or another if you have the development skills. You can create components to share, write your own nodes to expand functionality, incorporate other languages / platforms into KNIME, and there are even a few organizations out there that customize their own KNIME AP platform for internal use!

8 Likes

As you can tell, I am an unabashed KNIME fan! :joy:

The question I have is… Why isn’t an amazing open source tool like KNIME Analytics Platform ubiquitous in every industry?

9 Likes

Ivan:
KNIME and Alteryx are GUI-driven, and Base SAS is pure code (unless the SAS your are referring to is SAS-Enterprise Miner). I have not used Alteryx, but I have used SAS-EM a lot (and taught an online course with it. KNIME is far superior to SAS-EM if only for the very rich looping capabilities in KNIME. I used KNIME in my Data Science classes at the University of California, Irvine for 5 years before my retirement. My students loved it, particularly my international students. I found that the Hub examples (accessible in KNIME) and the Forum were rich sources of answers to all of my questions. The Forum doesn’t exist for Alteryx or SAS-EM; you are dependent on published documentation. The Forum provides very quick answers to specific questions, even those about nodes and workflows that don’t work right. It provides a large community of debuggers that you can plug into. KNIME also provides a wealth of training courses and materials. I have contributed a lot of my course material to KNIME.
Robert (Bob) Nisbet, PhD.
Instructor in the Data Science Certificate Program (Retired)
UC-Irvine

10 Likes

@Bob_Nisbet Tnx for additional explanation and insights Bob :slight_smile:
Ivan

2 Likes

Hi Robin,
I echo the other comments here that its the versatility and open nature of KNIME that is attractive. Both open source and commercial cheminformatics tools are available. There are also a lot of community contributions in the area of cheminformatics and image analysis available.
For me the main attractions of KNIME are reproducibility (you get a built in audit trail of what you have done), the ability to mix and match programming languages (people with different language skills can collaborate), the ability to read almost any datatype and database, the extensive data formatting and de-mangling tools and finally the BIRT report writer that helps automate report writing. There are also a lot of convenient ML tools and statistical and chemometrics functions are easily added with R scripting. KNIME has saved me years in manual data operations and reporting. I have not used Alteryx.

7 Likes

As has been mentioned, KNIME has since basically almost the beginning had extension for cheminformatics and later these got extended a lot by community contributions and by vendors of commercial software building nodes for KNIME. As far as I know Alteryx doesn’t offer “cheminformatics features”.

The real competitor to KNIME in this regard is Pipeline Pilot which is also a “GUI tool for pipelines” but initially made for Pharma. However the concepts are quiet different to KNIME and it is very costly just like Alteryx.

KNIME has the advanatge you can just start using it for free and from there it can organically grow until you potentially can profit from a knime server.

About image analysis, yes I’m pretty sure pharma uses it as it is often used to analyse microscope images like from cell cultures.

3 Likes

I’m using it for something completely different than most people, I feel. I pull data from applications and systems and compare the data to other databases and Active Directory, and use it as a tool to check security, and to prepare audits for compliance purposes. It got too tiring in Excel and SQL to continually have to redo the same steps when new data comes in.
I was presenting data to one of our leaders, and he said that the report needed to be redone as there had been changes in the past few hours. I don’t think he expected me to click a couple of buttons and have it magically refresh in front of him.

I wish I had known about it years ago.

7 Likes