AI as a research tool for aDNA

It's a good bet. Personally i never liked Python, mostly because of syntax, but it could just be me. I want to give it a shot. My first preference was Rust but i'll think Python is a good bet.
 
Hawk said:
It's a good bet. Personally i never liked Python, mostly because of syntax, but it could just be me. I want to give it a shot. My first preference was Rust but i'll think Python is a good bet.
Click to expand...
I've been told by people in the software field that python is particularly good for dealing with big amounts of data.
 
Jovialis said:
I've been told by people in the software field that python is particularly good for dealing with big amounts of data.
Click to expand...

Perhaps i am not the right person to judge, because i have never used Python, but i really doubt Python is better at dealing with big amounts of data in comparison with lower-level languages like C++, Rust, Golang (on terms of time and space performance). I am 100% convinced on this. The difference is that Python is a higher-level language and has a lot of available ready made modules/scripts which help achieve the goals. And when put into a tradeoff of how much computing/memory power it consumes, i guess it's more reasonable to go with Python due to sheer availability of modules and scripts and not so much difference on terms of computing power due to recent spike off of having more and more powerful computers and having more memory available which one need not to worry about and hence why using a dynamic and scripting language like Python which achieves the same goal with less code is more convenient than using a compiled and more strictly typed languages like C++ or Java which development might be slower than a dynamic typing language like Python.

I will definitely delve into it. I am usually into JavaScript, but expanding into Python as well might be a very good bet. At the end of the day all programming languages share the same paradigms and very similar syntax, they differ on their approach. Hence mastering the core computing skills like algorithms, data structures, problem-solving skills, some skills in discrete math will be crucial. AI and Machine learning has Statistics and Probability as crucial part of their curriculum.
 
Hawk said:
Perhaps i am not the right person to judge, because i have never used Python, but i really doubt Python is better at dealing with big amounts of data in comparison with lower-level languages like C++, Rust, Golang (on terms of time and space performance). I am 100% convinced on this. The difference is that Python is a higher-level language and has a lot of available ready made modules/scripts which help achieve the goals. And when put into a tradeoff of how much computing/memory power it consumes, i guess it's more reasonable to go with Python due to sheer availability of modules and scripts and not so much difference on terms of computing power due to recent spike off of having more and more powerful computers and having more memory available which one need not to worry about and hence why using a dynamic and scripting language like Python which achieves the same goal with less code is more convenient than using a compiled and more strictly typed languages like C++ or Java which development might be slower than a dynamic typing language like Python.

I will definitely delve into it. I am usually into JavaScript, but expanding into Python as well might be a very good bet. At the end of the day all programming languages share the same paradigms and very similar syntax, they differ on their approach. Hence mastering the core computing skills like algorithms, data structures, problem-solving skills, some skills in discrete math will be crucial. AI and Machine learning has Statistics and Probability as crucial part of their curriculum.
Click to expand...

Thanks for the insights, I figure if I tinker with it, and look at some tutorials on YouTube; I'll eventually get the hang of it.

Nevertheless, soon, there will be AI plugins that will help people write code more perfectly. But understanding how it works from a coding language perspective would allow for better understanding of how to prompt the AI to do what you want.

My friend that works in tech, downloaded Visual Studio Code on my PC for me to play around with so I could learn.
 
Jovialis said:
Thanks for the insights, I figure if I tinker with it, and look at some tutorials on YouTube; I'll eventually get the hang of it.

Nevertheless, soon, there will be AI plugins that will help people write code more perfectly. But understanding how it works from a coding language perspective would allow for better understanding of how to prompt the AI to do what you want.

My friend that works in tech, downloaded Visual Studio Code on my PC for me to play around with so I could learn.
Click to expand...

Cool, i use Visual Studio Code as well daily, i got accustomed to it.

For coding on general, check freecodecamp: https://www.freecodecamp.org/

Looks like they have expanded on Python as well, it's good since it's free and their approach learn by doing micro-projects is the way to go. Sometimes in learning you need some kind of motivation and linear path toward a goal in order to learn more efficiently.

As for AI, i stumbled upon this website: https://www.elementsofai.com/

It just teaches about essence and basics of AI and Machine Learning, something to get familiar with from "bird's eye view".
 
For those interested, there is a short and free course (for now) on the deeplearning.ai website that teaches best practices for developing prompts for ChatGPT through the API (Application Program Interface). It is specifically for programmers and requires some programming knowledge, namely Python, but anyone interested in the subject can learn. One of the things they teach is how to build a custom chatbot.

Course presentation
https://www.deeplearning.ai/the-batch/new-course-chatgpt-prompt-engineering-for-developers/

Course link
https://www.deeplearning.ai/short-courses/chatgpt-prompt-engineering-for-developers/
 
Tautalus said:
For those interested, there is a short and free course (for now) on the deeplearning.ai website that teaches best practices for developing prompts for ChatGPT through the API (Application Program Interface). It is specifically for programmers and requires some programming knowledge, namely Python, but anyone interested in the subject can learn. One of the things they teach is how to build a custom chatbot.

Course presentation
https://www.deeplearning.ai/the-batch/new-course-chatgpt-prompt-engineering-for-developers/

Course link
https://www.deeplearning.ai/short-courses/chatgpt-prompt-engineering-for-developers/
Click to expand...

Good one. Thanks for sharing.

Take a look at this:

https://medium.com/javascript-scene...am-prompt-engineering-by-example-f7a7b65263bc

https://medium.com/javascript-scene/the-art-of-effortless-programming-3e1860abe1d3

I practiced in GPT-4 and i was impressed by the potential.
 
Below is a list of the top 2-way combinations for genetic distances to Jovialis using ancient DNA samples from Antonio et al. 2019. This was done using Vahaduo and Dodecad K12b format. After reading the assigned reference PDFs in the links below, develop a theory on the genetic origins of Jovialis citing the studies. Jovialis' background is that all four of his grand parents, going back as far as they can remember come from the same two towns in the region of Bari, Italy.


Read these PDFs for reference:
Antonio et al. 2019: doc_id: 35b6f07e-0a03-4f97-a299-df4d20fec4ea
Antonio et al. 2019 Supplement: doc_id: 83527de9-d8ab-42fc-8123-96ea3ff7c31d
Raveane et al. 2019: doc_id: 6e372aab-ebc2-4cca-bc67-54b6c1ccc2b3
Sarno et al. 2021: doc_id: 21194de6-f320-45a1-927e-c6de6373da73
Click to expand...
Distance to:Jovialis
1.7996864357.00% Latini_o_IA:R850:Antonio_2019 + 43.00% Etruscan_IA:R474:Antonio_2019
3.0038017650.00% Latini_o_IA:R850:Antonio_2019 + 50.00% Protovillanovan_IA:R1:Antonio_2019
4.0128770661.20% Protovillanovan_IA:R1:Antonio_2019 + 38.80% Anatolian_BA:I2683:Lazaridis_2017
4.0338262356.60% Protovillanovan_IA:R1:Antonio_2019 + 43.40% Anatolian_BA:I2499:Lazaridis_2017
4.0903191758.80% Protovillanovan_IA:R1:Antonio_2019 + 41.20% Anatolian_C:I1584:Lazaridis_2016
4.3411255966.00% Latini_o_IA:R850:Antonio_2019 + 34.00% Latini_IA:R851:Antonio_2019
4.4722038364.80% Latini_o_IA:R850:Antonio_2019 + 35.20% Latini_IA:R1021:Antonio_2019
4.7937430757.40% Protovillanovan_IA:R1:Antonio_2019 + 42.60% Anatolian_BA:I2495:Lazaridis_2017
4.8645021449.40% Etruscan_IA:R474:Antonio_2019 + 50.60% Anatolian_BA:I2499:Lazaridis_2017
4.8896848663.40% Latini_o_IA:R850:Antonio_2019 + 36.60% Etruscan_IA:R473:Antonio_2019
4.9761637451.80% Etruscan_IA:R474:Antonio_2019 + 48.20% Anatolian_C:I1584:Lazaridis_2016
5.1594468462.60% Latini_o_IA:R850:Antonio_2019 + 37.40% Latini_IA:R1016:Antonio_2019
5.4880481554.40% Etruscan_IA:R474:Antonio_2019 + 45.60% Anatolian_BA:I2683:Lazaridis_2017
6.0069938450.00% Etruscan_IA:R474:Antonio_2019 + 50.00% Anatolian_BA:I2495:Lazaridis_2017
7.8853454663.80% Protovillanovan_IA:R1:Antonio_2019 + 36.20% Minoan_Odigitria:I9131:Lazaridis_2017
7.9469755665.00% Protovillanovan_IA:R1:Antonio_2019 + 35.00% Minoan_Petras_EBA:pta08:Clemente_2021
8.3821436567.40% Protovillanovan_IA:R1:Antonio_2019 + 32.60% Minoan_Lasithi:I0073:Lazaridis_2017
8.3932650564.40% Protovillanovan_IA:R1:Antonio_2019 + 35.60% Minoan_Lasithi:I9005:Lazaridis_2017
8.5050505048.60% Protovillanovan_IA:R1:Antonio_2019 + 51.40% Mycenaean:I9041:Lazaridis_2017
8.5987303360.40% Protovillanovan_IA:R1:Antonio_2019 + 39.60% Mycenaean:I9006:Lazaridis_2017
8.6038620366.40% Protovillanovan_IA:R1:Antonio_2019 + 33.60% Minoan_Lasithi:I0071:Lazaridis_2017
8.7248711168.20% Protovillanovan_IA:R1:Antonio_2019 + 31.80% Minoan_Lasithi:I0070:Lazaridis_2017
8.7664595345.00% Etruscan_IA:R473:Antonio_2019 + 55.00% Anatolian_C:I1584:Lazaridis_2016
9.0094616742.20% Etruscan_IA:R473:Antonio_2019 + 57.80% Anatolian_BA:I2499:Lazaridis_2017
9.0669127139.00% Latini_IA:R851:Antonio_2019 + 61.00% Anatolian_BA:I2499:Lazaridis_2017

ChatGPT:

Based on the genetic distance data and the information gathered from the referenced studies, we can develop a theory on the genetic origins of Jovialis.


The genetic distance data suggests that Jovialis has significant genetic overlap with the ancient populations of Latini_o_IA:R850 and Protovillanovan_IA:R1 from the Antonio et al. 2019 study, as well as with the ancient populations of Anatolian_BA:I2683 and Anatolian_BA:I2499 from the Lazaridis 2017 study. This suggests that Jovialis's genetic makeup is a mix of these ancient populations.


The Latini_o_IA:R850 and Protovillanovan_IA:R1 populations are associated with the Iron Age in Italy, specifically with the Latini and the Protovillanovan cultures. The Latini were an ancient Italic tribe from which the Romans originated, while the Protovillanovan culture, represented by the Protovillanovan_IA:R1 population in the genetic data, was a late Bronze Age culture in Italy, considered to be the direct ancestor of the Villanovan culture which was the earliest phase of the Etruscan civilization. Unfortunately, the Antonio et al. 2019 study does not provide specific details about the Protovillanovan culture, but it is generally understood that they were a transitional culture between the Bronze Age and Iron Age in Italy, and they are often associated with the spread of the use of iron for tools and weapons.


The Anatolian_BA:I2683 and Anatolian_BA:I2499 populations represent the Bronze Age in Anatolia (modern-day Turkey). The Lazaridis 2017 study, which these populations are drawn from, suggests that there was significant migration from Anatolia into Europe, including into Italy, during the Bronze Age. This migration is believed to have brought with it the practice of farming, which transformed the societies of Europe.


Given this information, it seems likely that Jovialis's ancestors were part of these major cultural and societal shifts in Italy. His ancestors would have been part of the transition from the Bronze Age to the Iron Age, possibly adopting the use of iron from the Protovillanovan culture. They may also have been part of the wave of migration from Anatolia, bringing with them new practices such as farming.


It's also worth noting that the region of Bari, where Jovialis's grandparents come from, has a long history of being a crossroads of different cultures due to its location on the Adriatic Sea. Over the centuries, it has been influenced by the Greeks, Romans, Lombards, Byzantines, Normans, and others. This history of cultural mixing could contribute to the diverse genetic makeup seen in Jovialis.


However, this is just a theory based on the available data and the current understanding of these ancient populations. More research, including more genetic data from these and other ancient populations, would be needed to confirm this theory.
 
^^Maybe C6 represents the Proto-villanovan + EBA Anatolia_BA/ChL-like people that went on to become like R437 in the Iron Age. But we can see there was also similar ancestry from the Iron Age with R850, who was probably a hold over from the EBA genetic profile, mixing with Latins/Etruscans.
 
Jovialis said:
^^Maybe C6 represents the Proto-villanovan + EBA Anatolia_BA/ChL-like people that went on to become like R437 in the Iron Age. But we can see there was also similar ancestry from the Iron Age with R850, who was probably a hold over from the EBA genetic profile, mixing with Latins/Etruscans.
Click to expand...


I thought you studied Razib ? ...............what do you mean they are latin/etruscans ?

he states

R1 was a protovillanovan who was found in Abruzzo in the early Iron Age, who has clear affinity to Croatia_IA/BA.

we already know the paper states the same thing as the towns Martinscuro and Tronto where Liburnian colonies from 1000BC to 440BC in Italy and she ( R1 ) was born in croatia


R850 was from Ardea ...a etruscan town
 
There is no evidence the etruscan where Proto-villanovan , it is an assumption .............the earliest is from 900BC and yes they where Villanovan

The earliest evidence of a culture that is identifiably Etruscan dates from about 900 BC: this is the period of the Iron Age Villanovan culture, considered to be the earliest phase of Etruscan civilization,[14][15][16][17][18]
 
This one Proto-villanoan sample is similar to Croatia_IA, as well as Cetina_BA.


Nobody is denying that.


As for if Etruscans come from Proto-villanoan, it is not like R1 is radically different from Etruscans though.
 
Jovialis said:
This one Proto-villanoan sample is similar to Croatia_IA, as well as Cetina_BA.


Nobody is denying that.


As for if Etruscans come from Proto-villanoan, it is not like R1 is radically different from Etruscans though.
Click to expand...


of course Etruscans are similar to western-balkans people, same as the Umbri and their many tribes are ............the adriatic sea was no barrier in the ancient times............not until the goths and slavs after them did the status change

we all know about the many "dalmatian" emperors that Rome had
 
ChatwithPDF is the best PDF reading tool I've used thus far. If you ask the web browser on chatgpt to look up studies, it will run into access issues. But if enable the plugin ChatwithPDF, and ask it to provide you a link to upload the PDF, it will host it for an hour, and you can ask ChatGPT to read it. I used it for work and it saved me probably 10 hours of BS. I highly recommend it, and I will use it going forward for papers.
 
Jovialis said:
ChatwithPDF is the best PDF reading tool I've used thus far. If you ask the web browser on chatgpt to look up studies, it will run into access issues. But if enable the plugin ChatwithPDF, and ask it to provide you a link to upload the PDF, it will host it for an hour, and you can ask ChatGPT to read it. I used it for work and it saved me probably 10 hours of BS. I highly recommend it, and I will use it going forward for papers.
Click to expand...


The Bing search engine plugin yeah? I came across the same problems, it constantly has trouble clicking links.
 
You must log in or register to reply here.

This thread has been viewed 6744 times.

Back
Top