Home

Welcome to PharmaKarma – my personal site on everything relating to data mining, drug design and a bit of pharma industry – all from the perspective of “@home”.  In addition, the premise is that pharma synthetic medication, from the dedicated scientists – but also “karma” – perspective, wants to and can do good.

Karma – means action, work or deed – it also refers to the spiritual principle of cause and effect where intent and actions of an individual (cause) influence the future of that individual (effect). Good intent and good deed contribute to good karma and future happiness (see e.g. Wikipedia). This is were many hard-working scientist come in – even you, the reader, who can do research (to a certain extent) from the comfort of home and thus add to a positive reputation of (future) medications!

Be it with the help of myself and consulting to you, or if you want to try/do it yourself – data mining has tremendous possibilities for doing it even for free and with simple tools available from your home!

 

Data Mining

Data Mining

This section is all about  data-mining. With data in this context I am referring to chemical structures and associated data from (mainly) public databases. In general I use the workflow creator “Knime” and combine it as required with Java, SQL and different APIs. See here for more details.

 

About

About myself

My name is Alexander Minidis, Ph.D. in organic chemistry, with a specialty in medicinal chemistry and a burning interest in data mining and linking IT with research. That, and working in general cross-functionally is something I found most intriguing in my working life so far. Lately, the hype around AI hasn’t gone past me either!

I studied (organic) chemistry in Basel, Switzerland and did my Ph.D with Professor Andreas Pfaltz in Basel and continued at the Max-Planck-Institute (MPI) f. Kohlenforschung in Mülheim, Germany. After a Post-Doc with Prof. Donna Blackmond at the MPI followed by a Post-Doc with  Prof. Jan Bäckvall in Stockholm, Sweden, I started my industry career at the chemistry CRO Syntagon, Södertälje (Sweden). Missing research, I went to AstraZeneca R&D and became a medicinal chemist. There I continued working as team- and project leader. Over the years I had the opportunity to “dabble” in all sorts of activities, be it synthesis in the lab, data-mining, designing new molecules, writing patent applications, due diligence of new projects or CROs. This, combined with my latter experience, gives me a rather unique background of someone who was worked actively (not only by reading or collaboration) with nearly all aspects of pre-clinical drug research.

After the closure of AstraZeneca research unit two companions and I founded a start-up, Evomedicon AB, a company within drug research and project management. One of the things we were working with was the experience we garnered from our AZ time where we implement lean into research. Interesting to note is, that lean and scrum were (still is) disliked by many researchers, though if you simply take away the naming, one will notice that for efficient workflows lean is more or less already in place. Thus, if adapted specifically for research, lean and scrum can work! Anyway, after 2 years we decided though to close the company and follow other avenues. I continued with research at the Karolinska Institute for some years, using my industry background in an academic setting in the group of Prof. Taipale. Though I required a lot of new know-how regarding biochemical methodologies unknown to me as chemist, the thing required most was data-mining and even programming.

This lead me to support Chemnotia AB in their request to develop a software for analysis of chemical synthesis pathways of larger proteins, an ideal scenario for acting as interface as both, software person and researcher in the same person and in collaboration with the customers.

Shortly after this, I joined Collaborative Drug Design (CDD) as sort of a technical expert for the European “time-zone”, supporting internal and external customers with the main product, CDD Vault, an online database for pharmaceutical drug development. The job included a wide variety of tasks, all from writing knowledgbase articles, tutorial videos, demo-ing the software and setting it up for customers, as well handling enhancement requests, bug-hunting, or giving presentations at conferences.

After an offer by Medivir, I workes for a short period as application scientist & data-analyst, supporting researchers with new implementations as well as support for existing software systems – until Medivir announced closure of their R&D portion just two months into this new position. Not waiting what will happen there next, I applied and was recruited by Rise, Process and Development where I am working as Chemist/Project Leader.

Finally, I have done some occasional consulting work, two of the main customers are Awametox and Chemnotia, whom I support(ed) with QA work or data-mining/programming. This concept of being able to support others from home has evolved into this website!

Further details regarding my CV can be found on my LinkedIn profile.

Disclaimer: The topics & opinions on this web page are my own and are not representative of any of the companies mentioned here, nor of any of my former employers. This blog and website is independent of PharmaKarma-Consulting (despite being the same physical person behind the consulting part).

Blog

Sweet, another publication! Machine Learning in Reaction predictions!

Well, sort-of-kind-of peer-reviewed. The publication in question went through several peer-review rounds before we decided (due to lack of time and resources) to go with ChemRxiv and leave it at that. But at least it’s out there. Here is the link: https://doi.org/10.26434/chemrxiv.12613214.v1 Machine Learning to Reduce Reaction Optimization Lead Time – Proof of Concept with …

Pubchem updates (affects Knime workflows)

I realize I haven’t been working a lot (at all….) the past year with the data-mining part vs Pubchem. There has been an update to the assay section of the database: https://pubchemblog.ncbi.nlm.nih.gov/2018/04/27/pubchem-bioassay-tools-to-be-replaced/ I will try and get back to this hopefully soon since this will (is) affect(ing) the Knime workflow as presented here elsewhere – …

Links

Some links of interest with regards to design & mining @home (non-exhaustive list):

Distributed/Volunteer Computing

Other Blogs: