Thanks to slashdot, I downloaded the TV show. Thanks slashdot.
On the other hand, after watching 15 minutes of the show, I realized it was complete crap. It was a bunch of pseudo science masquarading as something vaguely interesting. The writing was horrible, and the characters were completely unbelievable.
Funny. What you call sci-fi, I'd call fantasy. Sure, there are spaceships, lasers, and lightsabers. Ultimately, it's not about the technology - that just 'works' (see: magic).
Sci Fi and speculative fiction seem to have a much stronger grounding in reality. SF doesn't need the obvious cues of laser/robot/AI/nanotech/etc to hit the reader off that something is altered away from the universe we know (which we'd just call fiction.) SF could be about an alternate timeline with the exact same technology. It's still internally consistent, and there are reasons why things are the way they are. I'd even hesitantly call science fiction a subset of SF.
Star trek and star wars are both fantasy. ST is a fantasy/western, actually. Please don't sully SF by attributing either of those stories.
So, I looked at the resume. A few things: the 'both major platforms' might be written out with what the actual platforms are (does HR know what the two platforms are, or did your resume go in the trash?)
Also, you list jobs, but have no descriptions of what you did at them. So, do you have real life work experience? I have no idea, because you only list enough for me to call someone, but I'd have no idea what to ask them. You could have been a sanitation engineer at all those companies, for all I know.
Why can't we use our computers to scale images? I have a 720p projector (Sanyo Z2), and I scale my movies up to 1280x720 on the fly. You need a bit of processor power, but it's much cheaper than the scalers that are mentioned in your post. In addition, I'd assume that I can buy newer software packages and 'upgrade' my algorithms more easily than selling my scaler and buying a new one...
The trick with patenting SNPs is that there are 10,000,000 common ones, so that's a lot of money to spend on patents to 'cover' your disease. The number of true postive believable association studies using SNP data is still very low (we're up around ~10 or so now, which is far better than the 2-3 we had a few years ago.)
So, that's 10 SNPs to patent - except most of them were published papers comming out of acedemia, so they can't be patented. Now, if you can create a drug that acts to affect the changed protien confirmation (or transcription level, etc) of that mutation, then rock on. That's a hell of a lot of effort.
For reference: I do bioinformatics, I'm an author on one of the believable papers (PPARg/diabetes), and I am currently using HapMap data to research selection in the human genome.
The amount of effort neccesary to find a single SNP that effects a disease is massive, and the number of SNPs is high enough that 'covering your bases' for any disease would be prohibitively cost expensive. Hell, even covering all the missense SNPs would be expensive, and that misses out potentially interesting SNPs in silent regions (or my research, which shows certain regions of the genome that are non-coding, but under selection.)
And then, you can work on realistic physics, and AI. After that, you can work on better stories, non-linear plots, etc.
There's a million directions that technologies can improve in for games. I don't see the world ending when graphics get to be photorealisitc (and that's not going to happen any time soon anyway - compare Pixar level graphics to today's PC, we're many years away from having the processor power to do that in realtime.)
All that Irvine Welsh I've been reading pays off as I can seamlessly translate your last sentence. You wouldn't know of any dictionaries between that dialect and American English, would you? There are still a few words I wasn't able to understand from the context.
I spoke with the DBA, as well as our resident database architect. The DA was the guy who helped me shape a few querries in the not-as obvious methodology.
Also, if you want to do something interesting, like join 3 tables together, create a summary of those, then link to dimension tables from that summary - it's better to do the summary as an inner view. That way, you join on a smaller number of rows as you link to each dimension table. If you don't, then the summary step comes LAST, and you wind up doing joins on all the dimensions with the full rowset. It's a HUGE difference in performance (think: 15 seconds vs. 2 hours).
Things get interesting when you're working with huge data sets.
Or, I could continue developing my research apps, as I can do all the SQL work without issue.
I would be interested in creating large object graphs at will easily, as I do that as often as I do summaries. I do find that if I'm going to pull those graphs, sometimes writing queries in specific ways will get me vast improvements.
Oracle is annoying, in that sometimes two queries that look very similar can have different performance - sometimes 'forcing' the oracle query runner to pick a certain place to start earns you good performance gains.
This occurs often if you have a dimension table that you select one item from - say you select "bob" from the dimension table, and join to some other table. Oracle will think that's selective - even though the other table uses "bob" for 60% of its rows. You also join to that table on some other FK, which is much more selective, but doesn't look as selective to oracle, so it starts in the 'wrong' place with the query tree.
I'm curious as to how well hibernate performs under these areas. But, it sounds like it's well worth fooling around with, and teams that I work with/near have adopted it, so it's probably good to learn (even if I just put it on my resume and have another tech to say "been there".)
High performance? I'm curious, as I'm heavily thinking about shifting from handwritten SQL and the Spring framework to hibernate. I do scientific research, and pull large data sets (.1 - 5M rows) out of a database to operate on.
Often, without carefully tuning these querries, I'm getting horrible performance. With well tuned querries (and appropriate indexing), performace is acceptable. If hibernate was smart enough to say - build anonymous views then join on them when it was a benefit to performance - then I'd be interested. If hibernate was smart enough to aggregate multiple rows of data down to a single row to map onto my summary data objects, I'd be interested. If I'm simply pushing the problem off to another framework, then I'm not sure if it buys me anything over my current methodology.
So, how well does hibernate handle the really complex query work?
As a bioinformatics guy (see above post by me) who generally works in the field of population genetics, I often get to deal with large sets of patient data. We call 'em phenotypes, and we use them to distinguish between our 'cases' and controls, to stratify our populations.
That's how we do those association studies. True, we don't have the immediate goal of improving medical care, but we manage huge sets of data. Reference: I've got a data warehouse that has over 550M rows of data in it. That's essentially one data set.
Looking at people doing pop genetics may be close to that bridge you're talking about.
Bioinformatics was something that I "just got into". Really.
I had 5 years under my belt of lab work at MIT, and was learning programming again (I took AP comp-sci in highschool, and had decided to learn some programming for the hell of it with friends who were working in the industry.) There was call at work for me to automate some of the analysis that I needed to do.
Doing some simple tests like a TDT (yeah, I like population genetics) by hand took a long time, and was error prone. I used a bit of my programming knowledge to cook something up in a day to do the work for me.
My boss was pleased, and I soon had another relatively simple project to work on. True, most of these problems were 'parse a file, be an accountant, return a result', but it was fun and exciting to have problems that impacted my work day, and made life easier. The bonus was learning something of programming work, too.
This lead me to take a number of classes at local schools, and start reading a ton of books. A few years later, I was able to get my first full time informatics job (and was at that point a reasonably good scientist, so I was a 'two for one' kinda guy.)
This has lead to more jobs, more difficult projects, and a lot of great learning. Now, I write research projects dealing with selection, rules based frameworks for data analysis, data clustering, etc. Some projects are tools for scientists in my labs. Some projects are my own research.
I just 'got into it'. *shrug*. I don't know how common it is, but my co-worker learned bioinformatics the same way, and we seem to be pretty competent - we've both got papers in Nature/Nature Genetics under our belts, and we're collaborating to be co-first authors on a soon to be reviewed nature paper.
Spring is really a bunch of toolkits with one branding. I do all my database work with spring, and I have only one XML config file that tells my software what/where the database is. If I was running in a container, then I wouldn't need any XML (aside from the configuration stuff for the container, but that's a write once type deal) to connect to my DB and run all my querries, stored procs, etc.
I haven't yet jumped on the hibernate bandwagon yet, because I often do very complicated sql queries on large data sets, and tuning them for speed is important - and once you do that, it seems like you miss out on much of the power of hibernate, but you get all the bonus work of implementing hiberate. Spring is extremely easy to use, and makes my life a whole lot easier.
As someone with a reasonably strong backround in molecular biology and population genetics (I do informatics analysis work at the Broad institute of harvard and mit on population genetics issues), I don't think this has anything to do with the "hand of god":)
If RNA is associated with protiens or lipids, those have to still be present at the time the progeny is created, and passed into that. I'm not exactly sure that even if that is an explanation, that it would be universally true - I'm no plant biologist, but I wouldn't expect to see every protien expressed in the progeny immediately so that RNA could stick around. And, this RNA has to stick around from the grandparents to the grandchildren - that's a LONG time.
I still haven't read the 'real' paper yet (waiting to get my hands on the paper copy of nature), but it seems pretty unlikely that RNA is staying around that long. I'm not sure if they did the more obvious things (I would guess so) like checking for pseudogenes, etc. But those seem more likely.
I await with interest new data to see how this goes forward, but I'm not sure that they have any reasonable evidence for their current hypothesis. Again, the nature paper may shed some light on the topic of 'RNA stores', but that's the sort of thing (given how much expression work is being done) that I would have thought people would notice an extra stockpile of RNA lying around - and that it's levels would stay constant regardless of changing backround conditions.
Of course, I'm not sure if anyone's even done expression profiling on this model organism yet...
Hey yeah! That reminds me, I'll be submitting work to nature soon (in the next month), and since I'm not 1,2,4,5, I'm still somehow going to be first author.
Oh wait, I wrote all the analysis code. Weeee. I guess I qualify for "going to the meetings."
Seriously, I don't think all labs work that way. You'll have the PI get the last spot on the paper, and usually the person who did the most work get one of the first spots. Unless you work with a bunch of credit stealing a-holes, then it's time to work somewhere else.
There were no extra functional copies of the gene, but were there pseudo genes available with a large amount of sequence similarity? They probably already checked that by blasting their gene against the genome, but it would be worth checking that out at fairly low stringency to see if there were portions of the gene somewhere else, and that was being used Beats me to how it would be used, though - I can't imagine a molecular mechanism for my story that is even remotely plausable...
Since central dogma states that info flows DNA->RNA (except in odd cases with viruses), how does the RNA exist in the organism to correct the DNA? RNA degrades pretty quickly (otherwise, regulation of protien levels would be difficult), so it's not going to be around.
Hm, unless this happens at meiosis, and the parent's plant corrects it - but how would *it* know, since it has one copy of each allele?
There isn't by chance a pseudogene that is encoding almost the same info as this gene, and it's correcting off that?
I'm very interested in reading the actual paper, and starting to get to the bottom of this. Maybe my group at the Broad will talk about this in the next week or two (even though we do human genetics, it would be pretty cool to talk about.)
I've seen instances where people massage their data in pop gen papers as well. They're what we call 'fun' when we have occasional journal discussions. There's nothing like ripping the #$(*#$(* out of a paper that uses crappy statistics.
Usually, that manipulation is fairly obvious in the paper. I'd hope that the more reputible journals would learn to filter out more of this garbage over time.
All my statistical homies can now give me a shout out for Bonferroni correction, empircal p-value generation via permutation, etc. Conserative estimates on statistics make me feel all fuzzy, and I don't instantly feel like shouting out "Winner's Curse!"
--Jim (my last p-val was 4E-27. I can correct that all day. Comming to a journal near you soon, I hope!)
If you want to make really smooth ice cream, you need small ice crystals to do so. Try putting the ice cream mixture in the bottom of the fridge, until it is very cold (but above the freezing point.) Then put it in the ice cream maker. You'll get smaller ice crystals this way, and a smoother consistancy.
And yes, I learn some of my cooking techniques from Alton Brown (Good Eats, where i learned this trick.)
Slashdot Mirror
Oh good. Since I wanted to download microsoft office (not a game, music, or video) it's still legal!
Thanks to slashdot, I downloaded the TV show. Thanks slashdot.
On the other hand, after watching 15 minutes of the show, I realized it was complete crap. It was a bunch of pseudo science masquarading as something vaguely interesting. The writing was horrible, and the characters were completely unbelievable.
This show was shitcanned? No kidding.
I'd have done the same.
Funny. What you call sci-fi, I'd call fantasy. Sure, there are spaceships, lasers, and lightsabers. Ultimately, it's not about the technology - that just 'works' (see: magic).
Sci Fi and speculative fiction seem to have a much stronger grounding in reality. SF doesn't need the obvious cues of laser/robot/AI/nanotech/etc to hit the reader off that something is altered away from the universe we know (which we'd just call fiction.) SF could be about an alternate timeline with the exact same technology. It's still internally consistent, and there are reasons why things are the way they are. I'd even hesitantly call science fiction a subset of SF.
Star trek and star wars are both fantasy. ST is a fantasy/western, actually. Please don't sully SF by attributing either of those stories.
If CoH is the best of what's out there, then I'd just stop playing MMRPGs entirely.
Actually, I did (after playing CoH), and now only play some occasional games of guild wars.
So, I looked at the resume. A few things: the 'both major platforms' might be written out with what the actual platforms are (does HR know what the two platforms are, or did your resume go in the trash?)
Also, you list jobs, but have no descriptions of what you did at them. So, do you have real life work experience? I have no idea, because you only list enough for me to call someone, but I'd have no idea what to ask them. You could have been a sanitation engineer at all those companies, for all I know.
Why can't we use our computers to scale images? I have a 720p projector (Sanyo Z2), and I scale my movies up to 1280x720 on the fly. You need a bit of processor power, but it's much cheaper than the scalers that are mentioned in your post. In addition, I'd assume that I can buy newer software packages and 'upgrade' my algorithms more easily than selling my scaler and buying a new one...
Even in the future, people know Mary Chung's is good eats? That's excellent.
-Jim
(cambridge is my hood)
The trick with patenting SNPs is that there are 10,000,000 common ones, so that's a lot of money to spend on patents to 'cover' your disease. The number of true postive believable association studies using SNP data is still very low (we're up around ~10 or so now, which is far better than the 2-3 we had a few years ago.)
So, that's 10 SNPs to patent - except most of them were published papers comming out of acedemia, so they can't be patented. Now, if you can create a drug that acts to affect the changed protien confirmation (or transcription level, etc) of that mutation, then rock on. That's a hell of a lot of effort.
For reference: I do bioinformatics, I'm an author on one of the believable papers (PPARg/diabetes), and I am currently using HapMap data to research selection in the human genome.
The amount of effort neccesary to find a single SNP that effects a disease is massive, and the number of SNPs is high enough that 'covering your bases' for any disease would be prohibitively cost expensive. Hell, even covering all the missense SNPs would be expensive, and that misses out potentially interesting SNPs in silent regions (or my research, which shows certain regions of the genome that are non-coding, but under selection.)
And then, you can work on realistic physics, and AI. After that, you can work on better stories, non-linear plots, etc.
There's a million directions that technologies can improve in for games. I don't see the world ending when graphics get to be photorealisitc (and that's not going to happen any time soon anyway - compare Pixar level graphics to today's PC, we're many years away from having the processor power to do that in realtime.)
All that Irvine Welsh I've been reading pays off as I can seamlessly translate your last sentence. You wouldn't know of any dictionaries between that dialect and American English, would you? There are still a few words I wasn't able to understand from the context.
I spoke with the DBA, as well as our resident database architect. The DA was the guy who helped me shape a few querries in the not-as obvious methodology.
Also, if you want to do something interesting, like join 3 tables together, create a summary of those, then link to dimension tables from that summary - it's better to do the summary as an inner view. That way, you join on a smaller number of rows as you link to each dimension table. If you don't, then the summary step comes LAST, and you wind up doing joins on all the dimensions with the full rowset. It's a HUGE difference in performance (think: 15 seconds vs. 2 hours).
Things get interesting when you're working with huge data sets.
Or, I could continue developing my research apps, as I can do all the SQL work without issue.
I would be interested in creating large object graphs at will easily, as I do that as often as I do summaries. I do find that if I'm going to pull those graphs, sometimes writing queries in specific ways will get me vast improvements.
Oracle is annoying, in that sometimes two queries that look very similar can have different performance - sometimes 'forcing' the oracle query runner to pick a certain place to start earns you good performance gains.
This occurs often if you have a dimension table that you select one item from - say you select "bob" from the dimension table, and join to some other table. Oracle will think that's selective - even though the other table uses "bob" for 60% of its rows. You also join to that table on some other FK, which is much more selective, but doesn't look as selective to oracle, so it starts in the 'wrong' place with the query tree.
I'm curious as to how well hibernate performs under these areas. But, it sounds like it's well worth fooling around with, and teams that I work with/near have adopted it, so it's probably good to learn (even if I just put it on my resume and have another tech to say "been there".)
High performance? I'm curious, as I'm heavily thinking about shifting from handwritten SQL and the Spring framework to hibernate. I do scientific research, and pull large data sets (.1 - 5M rows) out of a database to operate on.
Often, without carefully tuning these querries, I'm getting horrible performance. With well tuned querries (and appropriate indexing), performace is acceptable. If hibernate was smart enough to say - build anonymous views then join on them when it was a benefit to performance - then I'd be interested. If hibernate was smart enough to aggregate multiple rows of data down to a single row to map onto my summary data objects, I'd be interested. If I'm simply pushing the problem off to another framework, then I'm not sure if it buys me anything over my current methodology.
So, how well does hibernate handle the really complex query work?
As a bioinformatics guy (see above post by me) who generally works in the field of population genetics, I often get to deal with large sets of patient data. We call 'em phenotypes, and we use them to distinguish between our 'cases' and controls, to stratify our populations.
That's how we do those association studies. True, we don't have the immediate goal of improving medical care, but we manage huge sets of data. Reference: I've got a data warehouse that has over 550M rows of data in it. That's essentially one data set.
Looking at people doing pop genetics may be close to that bridge you're talking about.
Bioinformatics was something that I "just got into". Really.
I had 5 years under my belt of lab work at MIT, and was learning programming again (I took AP comp-sci in highschool, and had decided to learn some programming for the hell of it with friends who were working in the industry.) There was call at work for me to automate some of the analysis that I needed to do.
Doing some simple tests like a TDT (yeah, I like population genetics) by hand took a long time, and was error prone. I used a bit of my programming knowledge to cook something up in a day to do the work for me.
My boss was pleased, and I soon had another relatively simple project to work on. True, most of these problems were 'parse a file, be an accountant, return a result', but it was fun and exciting to have problems that impacted my work day, and made life easier. The bonus was learning something of programming work, too.
This lead me to take a number of classes at local schools, and start reading a ton of books. A few years later, I was able to get my first full time informatics job (and was at that point a reasonably good scientist, so I was a 'two for one' kinda guy.)
This has lead to more jobs, more difficult projects, and a lot of great learning. Now, I write research projects dealing with selection, rules based frameworks for data analysis, data clustering, etc. Some projects are tools for scientists in my labs. Some projects are my own research.
I just 'got into it'. *shrug*. I don't know how common it is, but my co-worker learned bioinformatics the same way, and we seem to be pretty competent - we've both got papers in Nature/Nature Genetics under our belts, and we're collaborating to be co-first authors on a soon to be reviewed nature paper.
Spring is really a bunch of toolkits with one branding. I do all my database work with spring, and I have only one XML config file that tells my software what/where the database is. If I was running in a container, then I wouldn't need any XML (aside from the configuration stuff for the container, but that's a write once type deal) to connect to my DB and run all my querries, stored procs, etc.
I haven't yet jumped on the hibernate bandwagon yet, because I often do very complicated sql queries on large data sets, and tuning them for speed is important - and once you do that, it seems like you miss out on much of the power of hibernate, but you get all the bonus work of implementing hiberate. Spring is extremely easy to use, and makes my life a whole lot easier.
As someone with a reasonably strong backround in molecular biology and population genetics (I do informatics analysis work at the Broad institute of harvard and mit on population genetics issues), I don't think this has anything to do with the "hand of god" :)
If RNA is associated with protiens or lipids, those have to still be present at the time the progeny is created, and passed into that. I'm not exactly sure that even if that is an explanation, that it would be universally true - I'm no plant biologist, but I wouldn't expect to see every protien expressed in the progeny immediately so that RNA could stick around. And, this RNA has to stick around from the grandparents to the grandchildren - that's a LONG time.
I still haven't read the 'real' paper yet (waiting to get my hands on the paper copy of nature), but it seems pretty unlikely that RNA is staying around that long. I'm not sure if they did the more obvious things (I would guess so) like checking for pseudogenes, etc. But those seem more likely.
I await with interest new data to see how this goes forward, but I'm not sure that they have any reasonable evidence for their current hypothesis. Again, the nature paper may shed some light on the topic of 'RNA stores', but that's the sort of thing (given how much expression work is being done) that I would have thought people would notice an extra stockpile of RNA lying around - and that it's levels would stay constant regardless of changing backround conditions.
Of course, I'm not sure if anyone's even done expression profiling on this model organism yet...
Hey yeah! That reminds me, I'll be submitting work to nature soon (in the next month), and since I'm not 1,2,4,5, I'm still somehow going to be first author.
Oh wait, I wrote all the analysis code. Weeee. I guess I qualify for "going to the meetings."
Seriously, I don't think all labs work that way. You'll have the PI get the last spot on the paper, and usually the person who did the most work get one of the first spots. Unless you work with a bunch of credit stealing a-holes, then it's time to work somewhere else.
There were no extra functional copies of the gene, but were there pseudo genes available with a large amount of sequence similarity? They probably already checked that by blasting their gene against the genome, but it would be worth checking that out at fairly low stringency to see if there were portions of the gene somewhere else, and that was being used Beats me to how it would be used, though - I can't imagine a molecular mechanism for my story that is even remotely plausable...
Since central dogma states that info flows DNA->RNA (except in odd cases with viruses), how does the RNA exist in the organism to correct the DNA? RNA degrades pretty quickly (otherwise, regulation of protien levels would be difficult), so it's not going to be around.
Hm, unless this happens at meiosis, and the parent's plant corrects it - but how would *it* know, since it has one copy of each allele?
There isn't by chance a pseudogene that is encoding almost the same info as this gene, and it's correcting off that?
I'm very interested in reading the actual paper, and starting to get to the bottom of this. Maybe my group at the Broad will talk about this in the next week or two (even though we do human genetics, it would be pretty cool to talk about.)
Homey D. Clown (Damon Wayans) from In living color (circa 1990) called, they want their tag line back.
Talk about freakin' dated.
If you want to be a bit more modern (think: this millenium), please say "That's BULLCRAP!!!!"
Yes, all in caps, or you don't have RealUltimatePower!
I've seen instances where people massage their data in pop gen papers as well. They're what we call 'fun' when we have occasional journal discussions. There's nothing like ripping the #$(*#$(* out of a paper that uses crappy statistics.
Usually, that manipulation is fairly obvious in the paper. I'd hope that the more reputible journals would learn to filter out more of this garbage over time.
All my statistical homies can now give me a shout out for Bonferroni correction, empircal p-value generation via permutation, etc. Conserative estimates on statistics make me feel all fuzzy, and I don't instantly feel like shouting out "Winner's Curse!"
--Jim
(my last p-val was 4E-27. I can correct that all day. Comming to a journal near you soon, I hope!)
Biochem, molecular bio, biotechnology - there's a whole ream of degrees out there.
I think the particular one would depend on your interests, but the life sciences are a good bet for fun, interesting work. I know I love my job...:)
One thing to note:
If you want to make really smooth ice cream, you need small ice crystals to do so. Try putting the ice cream mixture in the bottom of the fridge, until it is very cold (but above the freezing point.) Then put it in the ice cream maker. You'll get smaller ice crystals this way, and a smoother consistancy.
And yes, I learn some of my cooking techniques from Alton Brown (Good Eats, where i learned this trick.)
What if people are lazy and sloppy, and rush their project out the door so they have more free time to play at work?
Maybe he only hires 'good' people, or uses this to weed out the lazy people.