4 Univ of Idaho Students Murdered, Bryan Kohberger Arrested, Moscow, Nov 2022 #88

Status
Not open for further replies.
  • #801
Yes, AT will pull out all the stops, that's all they can do because he is guilty. At least BK can't claim ineffectiveness of council, which is ultimately good for the case.

In the end this Defendant is going to be found guilty by a jury of his peers and either sentenced to the DP or LWOP. Of that, I have no doubt.

MOO
That's right, and imo, any argument AT presents is also another opportunity for the prosecution to strengthen their case. If AT finds a gap, let the prosecution fill it.
 
  • #802
Have you read the PCA? I mean really critically read it and considered what it says? I admit, I thought it was really strong at first. But then I went back and really paid close attention the the alleged "facts" in it and I was shocked because there is literally no evidence in the PCA that proves BK committed these murders. It's all supposition and guessing on the part of the writer. It's really quite appalling.
Respectfully, can you confirm your understanding of what a PCA is meant to accomplish? Because your arguments seem based on it being much more than it really is. Based on your expectations of the PCA, I don't think anyone could be arrested without video evidence and possibly a confession.
 
  • #803
My own opinion is that there is no way of "documenting" the mental processes behind IGG. It's simply impossible.

On the one hand, it's technically easy. Even if there had not been somehow named Kohberger in Othram's database (or whatever database it was), there would still be matches, because we are all related. Naturally, the higher the percentage of match (let's say 25%), then it's a certainty that the two people in question are related. That's because, aside from point mutations, everyone's DNA is sourced from just one place (their parents, then their grandparents, then great grandparents and so on). The mathematics (and SNP choice and chromosomal span) of this process is not difficult, easily accomplished by a computer. Is the computer "thinking"? What is there to report about what the computer does?

Many of our own human cognitive functions are easily replicated by machine; some are not. Obviously, a human could sit down with 600,000 data points on a long read out and try to personally verify that two different samples match and at what percentage/degree of confidence. That human had better find the same thing the computer does, though, because when it comes to comparing long (apparently random) lists of letters, computers are awesome.

Those of you who have taken biology know that one-half of these letters must mirror the other half - so the 1.2 million nucleotides in those 600,000 SNP's can in fact be halved. Super. Still quite a chore - and the repetitive nature of the task - make it more prone to human error.

Here's the type of chart that emerged from Idaho State Labs after putting the sample from the sheath into a DNA sequencer (for one in action, go to youtube and find Spencer Wells's The Journey of Man and skip to about 45 minutes, in the part where Wells is seeking one rare and specific SNP in India (an SNP that heretofore had only shown up in the Australian aborigine population - and it's my understanding that all of the Australians had that rare SNP).

1693668547966.png

This data is part of someone's genome. Make it way longer and that's what was sent to Othram, IMO. That's how DNA data is recorded, everywhere on Earth that I know of (because each letter stands for a chemical term - like guanine - and all biochemists use the same terms.

A machine then looked through their database (like the search function here on Websleuths) and found a significant match. Match means "is identical to" a certain segment of this data. Since Othram has chosen the 600,000 SNP's most associated with individual variation in 🤬🤬🤬🤬 sapiens, and since geneticists know the relative distance of two samples by statistical methods, this is quite effective at finding our relatives. I have 5000 relatives on 23andme. Some of them are descendants from 6xGreat grandparents, of course. But the system is sensitive enough to detect that. One doesn't even need to write anything down - the computer does the "writing," and using pen and pencil would be silly and impractical. The computer does the "thinking" and "recording."

Kind of like the chambers on a slot machine clicking into place - with just 4 symbols (ACTG) repeated 600,000 times in a specific sequence (your own DNA is as specific as it gets - unless you are an identical twin).

If the Defense was asking the Kohberger family to reveal whether Bryan has an heretofore unknown identical twin, and asking them to produce evidence that such a twin did NOT exist - it would be similar. One can MATCH things (children can do it), but to say what is NOT present in a pattern is quite difficult - esp. if the thing NOT there doesn't exist in the first place. In the case of DNA, it would be obvious if something other than ATCG were there (but I've never ever seen a computer make a mistake and put a wrong letter in - as they are only given FOUR to work with).

IMO.
 
  • #804
Regardless of the outcome, the FBI's paper trail should be made available to the defense. Not by Judge Judge compelling Latah County to produce it (because it's not theirs to produce) but by the FBI directly. Then it will be up to the defense to see if there is anything in it (or omitted from it) that would suggest abuse of process. At the very least that should happen IMO, because it's in the DOJ's rules (which I'll link below):
"If a suspect is arrested and charged with a criminal offense after an FGG profile has been entered into one or more DTC services, the investigative agency shall make a prompt formal request that all FGG profiles and associated account information and data held by any such service be removed from its records and provided directly to the investigative agency. The investigative agency shall document its request and compliance by the DTC service(s). All FGG profiles, account information, and data shall be retained by the investigative agency for potential use during prosecution and subsequent judicial proceedings."

"If a suspect is arrested and charged with a criminal offense after an FGG profile has been entered into an open-data personal genomics DNA database, the investigative agency shall promptly remove the FGG profile and all associated account information and data from the database. The investigative agency shall document the removal of this information and data. It shall be retained by the investigative agency for potential use during prosecution and subsequent judicial proceedings."


It seems reading all of the cases quoted upthread (and others) that there is no chance of a 4th Amendment argument to privacy on the IGG/FGGS. That appears to be settled unless BK wants to be the one to test this at the US Supreme Court (IGG hasn't yet been explicitly covered by the third-party doctrine to Constitutional rights, it has never been tested at the USSC, and maybe he wants to be the one to try it, who knows).

 
  • #805
The only other DNA thing might be the SODDI defense that the "three unknown male DNA samples" points to. This will likely be the defense's main attack at trial - IMO - if they can't have the entire DNA piece thrown out.

I don't think they will get it thrown out unless there is an abuse of process made clear in the FBI's notes, which for all we know AT may already have. The earlier testimony that stated that agencies often can get around opt-in/opt-out might be more telling than we assume, and would certainly put the agency foul of the DOJ rules linked in the post just before this.
 
  • #806
I understand why people might want to believe that but it isn't true.

The Department of Justice's rules on use of IGG uses the term "shall document" six times, and that's just the publicly available version (not an officer's handbook/departmental rules).

This sentence stood out to me:

"All FGG profiles, account information, and data shall be retained by the investigative agency for potential use during prosecution and subsequent judicial proceedings."

Those are the rules that FBI agents doing this work would have to follow. It's not a couple of agents huddled round a laptop with a Othram portal open and no paper-trail being made.

The 'shall document' rule is for those steps that can be documented. And those notes have been handed over already.
 
  • #807
The 'shall document' rule is for those steps that can be documented. And those notes have been handed over already.
I don't believe they have. Not the IGG ones, not the notes about how the FBI actually used the databases that they use and profiles they use to conduct FGGS/IGG investigations.
 
  • #808
Regardless of the outcome, the FBI's paper trail should be made available to the defense. Not by Judge Judge compelling Latah County to produce it (because it's not theirs to produce) but by the FBI directly. Then it will be up to the defense to see if there is anything in it (or omitted from it) that would suggest abuse of process. At the very least that should happen IMO, because it's in the DOJ's rules (which I'll link below):
"If a suspect is arrested and charged with a criminal offense after an FGG profile has been entered into one or more DTC services, the investigative agency shall make a prompt formal request that all FGG profiles and associated account information and data held by any such service be removed from its records and provided directly to the investigative agency. The investigative agency shall document its request and compliance by the DTC service(s). All FGG profiles, account information, and data shall be retained by the investigative agency for potential use during prosecution and subsequent judicial proceedings."

"If a suspect is arrested and charged with a criminal offense after an FGG profile has been entered into an open-data personal genomics DNA database, the investigative agency shall promptly remove the FGG profile and all associated account information and data from the database. The investigative agency shall document the removal of this information and data. It shall be retained by the investigative agency for potential use during prosecution and subsequent judicial proceedings."


It seems reading all of the cases quoted upthread (and others) that there is no chance of a 4th Amendment argument to privacy on the IGG/FGGS. That appears to be settled unless BK wants to be the one to test this at the US Supreme Court (IGG hasn't yet been explicitly covered by the third-party doctrine to Constitutional rights, it has never been tested at the USSC, and maybe he wants to be the one to try it, who knows).


I suspect the "paper trail" looks very similar to the picture I just posted above. It has been turned over. What other paper trail would there be? DNA doesn't leave a paper trail. The computer can print out the letters and the database can give a percentage of a match - I suppose someone could have jotted it down, but if they didn't, it's not that hard to keep in one's head.

BK might want to test things at the Supreme Court, for sure - but that's going to be way down the road and at the rate this whole thing is moving, there will be precedent-setting case law from many places before he ever gets there.

IGG is one of hundreds of thousands, perhaps millions, of data techniques that the Supreme Court has never looked at. It's kind of like making a bar graph out of an Excel spreadsheet. No one seems to be upset with charts and graphs. That's what a genealogy is (a chart). Created, in this case, by a computer.

The computer came up with some names (it doesn't have a thought process, IMO). The names were given to the Moscow PD and their investigation. They wrote them down. That's been produced. The DNA match was with a person named Kohberger in the IGG database.

I'm not sure how anyone can provide a paper trail of a process like this (do we all need to log keystrokes? That's not going to be mandated by any higher court, any time soon, IMO). At any rate the FBI has clear instructions on how to get one of their investigators to testify (orally) at trial. And the Defense has declined to try that procedure - instead asking the State to do it (but the State is under no obligation to do it). Wild Good Chase - but, fodder for the news until something else happens.

IMO. I do think the Othram-Kohberger (if that's the service) has some privacy rights, btw. And if a federal court wants to break into those rights - that's what has to happen next (via the FBI subpoena route, obviously). I figure that will extend this case somewhat indefinitely. While Kohberger continues to sit in jail.

It will be interesting to see if the Defense can even compel the FBI investigator in question to testify at trial (and I totally understand their reluctance to try to get that person to speak under oath - on both sides),

All my opinion.
 
  • #809
Second, for some reason, Imo, you may be continuing to conflate IGG with the reports related to the dna that are already in discovery and available to the defense. Namely and in my own words from reading the motions in this case, Reports that:
a)show exactly how the dna was extracted from the snap button;
b) show exactly how it was STR profiled producing what's known as as a "forensic profile/sample";
c) show exactly how it was put into CODIS and no match identified;
d)show exactly how the forensic dna (suspect) sample/profile was kept safe in the Idaho State lab;
e)show exactly the test performed by a private lab to produce an snp profile from the "forensic profile";
f) show exactly how the forensic sample was still safely stored in the Idaho State lab;
g) show exactly how the forensic sample was tested against a legally gathered trash sample (from the trash of BK's family home) on 27th Dec in the Idaho State Lab;
h) show exactly how that test showed that the donor of the trash sample was 99.9998%(approx from memory ) likely to be the father of the donor of the suspect forensic sample on the sheath stored safely in the Idaho State Lab;
i) show exactly how a buccal swab was taken from BK upon his arrest, and exactly how the dna from that buccal swap was a direct match (in the range of octillions of probabilty) with the forensic suspect sample safely stored in the Idaho State lab. Moo
<snipped for focus>

Wondering where you would place the FBI's IGG search on this list. Was it before the retrieval of Kohberger's father's DNA from the trash? And, if so, would that be significant in some way to the defense's strategy?
 
  • #810
I suspect the "paper trail" looks very similar to the picture I just posted above. It has been turned over. What other paper trail would there be? DNA doesn't leave a paper trail. The computer can print out the letters and the database can give a percentage of a match - I suppose someone could have jotted it down, but if they didn't, it's not that hard to keep in one's head.

BK might want to test things at the Supreme Court, for sure - but that's going to be way down the road and at the rate this whole thing is moving, there will be precedent-setting case law from many places before he ever gets there.

IGG is one of hundreds of thousands, perhaps millions, of data techniques that the Supreme Court has never looked at. It's kind of like making a bar graph out of an Excel spreadsheet. No one seems to be upset with charts and graphs. That's what a genealogy is (a chart). Created, in this case, by a computer.

The computer came up with some names (it doesn't have a thought process, IMO). The names were given to the Moscow PD and their investigation. They wrote them down. That's been produced. The DNA match was with a person named Kohberger in the IGG database.

I'm not sure how anyone can provide a paper trail of a process like this (do we all need to log keystrokes? That's not going to be mandated by any higher court, any time soon, IMO). At any rate the FBI has clear instructions on how to get one of their investigators to testify (orally) at trial. And the Defense has declined to try that procedure - instead asking the State to do it (but the State is under no obligation to do it). Wild Good Chase - but, fodder for the news until something else happens.

IMO. I do think the Othram-Kohberger (if that's the service) has some privacy rights, btw. And if a federal court wants to break into those rights - that's what has to happen next (via the FBI subpoena route, obviously). I figure that will extend this case somewhat indefinitely. While Kohberger continues to sit in jail.

It will be interesting to see if the Defense can even compel the FBI investigator in question to testify at trial (and I totally understand their reluctance to try to get that person to speak under oath - on both sides),

All my opinion.
Please then, why does the Department of Justice instruct FBI agents to retain documentation of its uses of FGGS/IGG services for trial?

It's not just jotted down if the agent feels like doing that, it is a rule they must follow to do their work:

"All FGG profiles, account information, and data shall be retained by the investigative agency for potential use during prosecution and subsequent judicial proceedings."

That means they must - by order of their employer which happens to be the DOJ - document exactly how they have used an IGG database in every single investigation they do. It's already mandated.

 
  • #811
My own opinion is that there is no way of "documenting" the mental processes behind IGG. It's simply impossible.

On the one hand, it's technically easy. Even if there had not been somehow named Kohberger in Othram's database (or whatever database it was), there would still be matches, because we are all related. Naturally, the higher the percentage of match (let's say 25%), then it's a certainty that the two people in question are related. That's because, aside from point mutations, everyone's DNA is sourced from just one place (their parents, then their grandparents, then great grandparents and so on). The mathematics (and SNP choice and chromosomal span) of this process is not difficult, easily accomplished by a computer. Is the computer "thinking"? What is there to report about what the computer does?

Many of our own human cognitive functions are easily replicated by machine; some are not. Obviously, a human could sit down with 600,000 data points on a long read out and try to personally verify that two different samples match and at what percentage/degree of confidence. That human had better find the same thing the computer does, though, because when it comes to comparing long (apparently random) lists of letters, computers are awesome.

Those of you who have taken biology know that one-half of these letters must mirror the other half - so the 1.2 million nucleotides in those 600,000 SNP's can in fact be halved. Super. Still quite a chore - and the repetitive nature of the task - make it more prone to human error.

Here's the type of chart that emerged from Idaho State Labs after putting the sample from the sheath into a DNA sequencer (for one in action, go to youtube and find Spencer Wells's The Journey of Man and skip to about 45 minutes, in the part where Wells is seeking one rare and specific SNP in India (an SNP that heretofore had only shown up in the Australian aborigine population - and it's my understanding that all of the Australians had that rare SNP).

View attachment 444356

This data is part of someone's genome. Make it way longer and that's what was sent to Othram, IMO. That's how DNA data is recorded, everywhere on Earth that I know of (because each letter stands for a chemical term - like guanine - and all biochemists use the same terms.

A machine then looked through their database (like the search function here on Websleuths) and found a significant match. Match means "is identical to" a certain segment of this data. Since Othram has chosen the 600,000 SNP's most associated with individual variation in *advertiser censored* sapiens, and since geneticists know the relative distance of two samples by statistical methods, this is quite effective at finding our relatives. I have 5000 relatives on 23andme. Some of them are descendants from 6xGreat grandparents, of course. But the system is sensitive enough to detect that. One doesn't even need to write anything down - the computer does the "writing," and using pen and pencil would be silly and impractical. The computer does the "thinking" and "recording."

Kind of like the chambers on a slot machine clicking into place - with just 4 symbols (ACTG) repeated 600,000 times in a specific sequence (your own DNA is as specific as it gets - unless you are an identical twin).

If the Defense was asking the Kohberger family to reveal whether Bryan has an heretofore unknown identical twin, and asking them to produce evidence that such a twin did NOT exist - it would be similar. One can MATCH things (children can do it), but to say what is NOT present in a pattern is quite difficult - esp. if the thing NOT there doesn't exist in the first place. In the case of DNA, it would be obvious if something other than ATCG were there (but I've never ever seen a computer make a mistake and put a wrong letter in - as they are only given FOUR to work with).

IMO.


Your expertise is so much appreciated.
thank you for the time you take to explain it so magnificently to non scientists, and anti scientists too.
 
  • #812
(but I've never ever seen a computer make a mistake and put a wrong letter in - as they are only given FOUR to work with).
There is such a thing as an intermittent software fault where database software - and that's what this is - occasionally makes the wrong input resulting in an incorrect result.

"An intermittent fault in computer software is a malfunction of a software program that occurs at intervals, usually irregular, while the software functions normally at other times."

"Avizienis et al. (2004) defines intermittent faults as the union of (a) elusive permanent faults, i.e. faults that manifest themselves conditionally, with their activation conditions depending on complex combinations of internal state and external requests, that occur rarely and can be very difficult to reproduce and (b) transient faults, which includes physical faults (i.e. faults associated with the hardware) as well as interaction faults, stemming from reciprocal actions with external systems. "

In my experience 100% of software programs have these intermittent faults or bugs, some more prominently than others. I have seen software that shows these bugs nearly every time you use it. I have even met users who believed it was "normal" and ignored the fact that the results were always wrong when the bug occurred and therefore made bad decisions based on bad results. It is harder to catch in software that shows these faults much more rarely, perhaps 1 in a random number of uses. These faults run the gamut from always to sometimes. Obviously they are more difficult to detect and prove if the fault is more rare. However there are a number of ways to detect the source of the problem.

You might work with the same software for many years and never notice a fault or it may occur and you recognize it. This is why replication of the process by an outside unbiased party is an imperative to prove that the result is accurate. As things stand in this case, we cannot and should not EVER trust that the IGG is accurate until we know that the result has been replicated by some unbiased outside party and that starts with documentation of the process initially used to get the IGG.

The problem here is not about trusting the DNA, it's about making certain that the software worked correctly.

10ofRods, from your perspective, have you ever had someone say that their DNA results doesn't show something they thought they were? For example someone who believed they were Indigenous American and it shows they have 0% Indigenous American DNA? I'm using the example, because I have a friend whose DNA results showed she was 0% Indigenous American, but her own sister is 32% Indigenous American. Their parents DNA results also show they both have Indigenous American DNA (a higher level - I don't remember quite what but maybe 60%) - all results for the daughters and parents came from the same DNA company, both women have the same mother and father. Can you explain these results to me? My friend is most perplexed as am I.
 
  • #813
<snipped for focus>

Wondering where you would place the FBI's IGG search on this list. Was it before the retrieval of Kohberger's father's DNA from the trash? And, if so, would that be significant in some way to the defense's strategy?
The name in the IGG database matched against the sheath. With the name Kohberger coming from that match

Then they went through the trash. Found the father of the sheath DNA person/suspect.
 
  • #814
There is such a thing as an intermittent software fault where database software - and that's what this is - occasionally makes the wrong input resulting in an incorrect result.

"An intermittent fault in computer software is a malfunction of a software program that occurs at intervals, usually irregular, while the software functions normally at other times."

"Avizienis et al. (2004) defines intermittent faults as the union of (a) elusive permanent faults, i.e. faults that manifest themselves conditionally, with their activation conditions depending on complex combinations of internal state and external requests, that occur rarely and can be very difficult to reproduce and (b) transient faults, which includes physical faults (i.e. faults associated with the hardware) as well as interaction faults, stemming from reciprocal actions with external systems. "

In my experience 100% of software programs have these intermittent faults or bugs, some more prominently than others. I have seen software that shows these bugs nearly every time you use it. I have even met users who believed it was "normal" and ignored the fact that the results were always wrong when the bug occurred and therefore made bad decisions based on bad results. It is harder to catch in software that shows these faults much more rarely, perhaps 1 in a random number of uses. These faults run the gamut from always to sometimes. Obviously they are more difficult to detect and prove if the fault is more rare. However there are a number of ways to detect the source of the problem.

You might work with the same software for many years and never notice a fault or it may occur and you recognize it. This is why replication of the process by an outside unbiased party is an imperative to prove that the result is accurate. As things stand in this case, we cannot and should not EVER trust that the IGG is accurate until we know that the result has been replicated by some unbiased outside party and that starts with documentation of the process initially used to get the IGG.

The problem here is not about trusting the DNA, it's about making certain that the software worked correctly.

But NONE of the above is relevant to this case. The IGG process was verified, in the end, to be correct. That was, IN FACT, BK's DNA on the sheath. So software glitches did not happen during the IGG process.

The state is trying to figure out if the defendant stabbed 4 students to death. The state is not concerned about 'making certain the IGG software worked without glitches.' That is not the concern because there was a wholly separate scientific process which correctly identified this defendant as 100% match to that unknown DNA--- Without using any of the IGG software.

AT is trying to send this court on a wild goose chase. IMO
 
  • #815
But NONE of the above is relevant to this case. The IGG process was verified, in the end, to be correct. That was, IN FACT, BK's DNA on the sheath. So software glitches did not happen during the IGG process.

The state is trying to figure out if the defendant stabbed 4 students to death. The state is not concerned about 'making certain the IGG software worked without glitches.' That is not the concern because there was a wholly separate scientific process which correctly identified this defendant as 100% match to that unknown DNA--- Without using any of the IGG software.

AT is trying to send this court on a wild goose chase. IMO

To be honest, she sent all of us on a wild goose chase. If it pointed to BK and ultimately his father's DNA did NOT match they would go look for the "real" murderer. But it DiID match, so the use of it reinforces the software is accurate.

I feel like we are going in circles for absolutely no reason (argument for arguments sake) just because it's called into question. We don't have to take that bait. Frankly, there are more interesting aspects to this case than the IGG suspect identifying data that was reinforced by the DNA cheek swab.

The courts may need to decide if we are truly allowed to use it. That remains to be seen. So arguing repeatedly about the use of it seems moot. JMOO
 
  • #816
Have you read the PCA? I mean really critically read it and considered what it says? I admit, I thought it was really strong at first. But then I went back and really paid close attention the the alleged "facts" in it and I was shocked because there is literally no evidence in the PCA that proves BK committed these murders. It's all supposition and guessing on the part of the writer. It's really quite appalling. The white car disappears in Pullman only to magically reappear in a residential neighborhood on the far side of Moscow from Pullman, begging the question of if it is even the same car? At first there was a white Nissan Altima, then, there were 22,000 2011-2013 white Elantra's that LE were searching through, then they changed the date range to include 2015 - an estimated total of 33,000 Elantras. The cellphone pings that look so damning at first ONLY mean his car was within 25 miles of the cellphone tower that 1122 King Rd uses. The night of the murders, his cellphone goes off in Pullman and comes on hours later near Genesee. As far as I can tell, the vehicle may have been down there during the murders. The description of the man in black has him at 5'10" when it is obvious BK is considerably taller than that. The only evidence they really have is his DNA on a snap on the sheath. No other DNA of his in the rooms where the murders were committed. There is literally nothing in the PCA proves he was even in Moscow that night, much less that he is a murderer. And if it is touch DNA on the sheath, he doesn't even have to have ever touched the sheath to get his DNA on it. It could have been, he worked out at the gym and some guy used a machine or weights BK used and transferred the DNA onto the sheath or BK shook hands with someone or used the same doorknob or touchpad at the store. Now I REALLY hope that the prosecution has more evidence than what was in the PCA and I'm waiting until the trial, if there is one, to render judgement. But I'm really quite concerned about this entire situation.

[BBM] Where are you getting this information? (or, who is telling you this?) TIA

"all supposition and guessing on the part of the writer." (They can't swear out an arrest warrant on a hunch imo. It won't issue. It's why they wait. jmo)

'The cellphone pings ... ONLY mean his car was within 25 miles of the cellphone tower that 1122 King Rd uses."

 
  • #817
My own opinion is that there is no way of "documenting" the mental processes behind IGG. It's simply impossible.

On the one hand, it's technically easy. Even if there had not been somehow named Kohberger in Othram's database (or whatever database it was), there would still be matches, because we are all related. Naturally, the higher the percentage of match (let's say 25%), then it's a certainty that the two people in question are related. That's because, aside from point mutations, everyone's DNA is sourced from just one place (their parents, then their grandparents, then great grandparents and so on). The mathematics (and SNP choice and chromosomal span) of this process is not difficult, easily accomplished by a computer. Is the computer "thinking"? What is there to report about what the computer does?

Many of our own human cognitive functions are easily replicated by machine; some are not. Obviously, a human could sit down with 600,000 data points on a long read out and try to personally verify that two different samples match and at what percentage/degree of confidence. That human had better find the same thing the computer does, though, because when it comes to comparing long (apparently random) lists of letters, computers are awesome.

Those of you who have taken biology know that one-half of these letters must mirror the other half - so the 1.2 million nucleotides in those 600,000 SNP's can in fact be halved. Super. Still quite a chore - and the repetitive nature of the task - make it more prone to human error.

Here's the type of chart that emerged from Idaho State Labs after putting the sample from the sheath into a DNA sequencer (for one in action, go to youtube and find Spencer Wells's The Journey of Man and skip to about 45 minutes, in the part where Wells is seeking one rare and specific SNP in India (an SNP that heretofore had only shown up in the Australian aborigine population - and it's my understanding that all of the Australians had that rare SNP).

View attachment 444356

This data is part of someone's genome. Make it way longer and that's what was sent to Othram, IMO. That's how DNA data is recorded, everywhere on Earth that I know of (because each letter stands for a chemical term - like guanine - and all biochemists use the same terms.

A machine then looked through their database (like the search function here on Websleuths) and found a significant match. Match means "is identical to" a certain segment of this data. Since Othram has chosen the 600,000 SNP's most associated with individual variation in *advertiser censored* sapiens, and since geneticists know the relative distance of two samples by statistical methods, this is quite effective at finding our relatives. I have 5000 relatives on 23andme. Some of them are descendants from 6xGreat grandparents, of course. But the system is sensitive enough to detect that. One doesn't even need to write anything down - the computer does the "writing," and using pen and pencil would be silly and impractical. The computer does the "thinking" and "recording."

Kind of like the chambers on a slot machine clicking into place - with just 4 symbols (ACTG) repeated 600,000 times in a specific sequence (your own DNA is as specific as it gets - unless you are an identical twin).

If the Defense was asking the Kohberger family to reveal whether Bryan has an heretofore unknown identical twin, and asking them to produce evidence that such a twin did NOT exist - it would be similar. One can MATCH things (children can do it), but to say what is NOT present in a pattern is quite difficult - esp. if the thing NOT there doesn't exist in the first place. In the case of DNA, it would be obvious if something other than ATCG were there (but I've never ever seen a computer make a mistake and put a wrong letter in - as they are only given FOUR to work with).

IMO.

OUTSTANDING!

jmo
 
  • #818
There is such a thing as an intermittent software fault where database software - and that's what this is - occasionally makes the wrong input resulting in an incorrect result.

"An intermittent fault in computer software is a malfunction of a software program that occurs at intervals, usually irregular, while the software functions normally at other times."

"Avizienis et al. (2004) defines intermittent faults as the union of (a) elusive permanent faults, i.e. faults that manifest themselves conditionally, with their activation conditions depending on complex combinations of internal state and external requests, that occur rarely and can be very difficult to reproduce and (b) transient faults, which includes physical faults (i.e. faults associated with the hardware) as well as interaction faults, stemming from reciprocal actions with external systems. "

In my experience 100% of software programs have these intermittent faults or bugs, some more prominently than others. I have seen software that shows these bugs nearly every time you use it. I have even met users who believed it was "normal" and ignored the fact that the results were always wrong when the bug occurred and therefore made bad decisions based on bad results. It is harder to catch in software that shows these faults much more rarely, perhaps 1 in a random number of uses. These faults run the gamut from always to sometimes. Obviously they are more difficult to detect and prove if the fault is more rare. However there are a number of ways to detect the source of the problem.

You might work with the same software for many years and never notice a fault or it may occur and you recognize it. This is why replication of the process by an outside unbiased party is an imperative to prove that the result is accurate. As things stand in this case, we cannot and should not EVER trust that the IGG is accurate until we know that the result has been replicated by some unbiased outside party and that starts with documentation of the process initially used to get the IGG.

The problem here is not about trusting the DNA, it's about making certain that the software worked correctly.

10ofRods, from your perspective, have you ever had someone say that their DNA results doesn't show something they thought they were? For example someone who believed they were Indigenous American and it shows they have 0% Indigenous American DNA? I'm using the example, because I have a friend whose DNA results showed she was 0% Indigenous American, but her own sister is 32% Indigenous American. Their parents DNA results also show they both have Indigenous American DNA (a higher level - I don't remember quite what but maybe 60%) - all results for the daughters and parents came from the same DNA company, both women have the same mother and father. Can you explain these results to me? My friend is most perplexed as am I.

Quick answer to the last question (and a slightly longer answer below).

While all offspring (in sexually reproducing creatures) have two parents (and usually four grandparents), they do not inherit equally from their grandparents. It's possible for siblings to have very different DNA. While most of us cluster around the mean, some of us don't.

Rest of the answer: don't know what service we're speaking of, but all of the ones I know about have revised their SNP cards for "indigenous people" many times. My own numbers have bounced around. Just in the last year, there have been two revisions (with lots of people dropping to 0% indigenous due to controversy over certain markers - like Taino).

More answer: sometimes people's parents are not who they think they are. It's been a finding in anthropology ever since this DNA-testing thing became inexpensive. There are lots of families (depends on which culture we're speaking of) where the children believe the marital couple are their parents - but DNA shows otherwise.

Getting tested for DNA/ancestry can open a big can of worms. It's not my place to suggest that you go tell them, "Hey, maybe you're not both the children of this couple," but...well, that's definitely a possibility. If both parents are 60% indigenous (that's one of the highest amounts I've ever heard come out of Ancestry or 23andme), then perhaps one of the children is the child of someone else. OTOH, did you actually see the results? I have convos with people frequently where their lists of what they are have shifted over time - because they are misremembering what they read in the report.

At any rate, most of my own results did not add up with what I was told I was. Even my bio mom was ill-informed about my bio-dad's family - but it's quite clear that my results are accurate and I have an aunt who did the pedigree for her family and already had the documentation. However, I'm well aware that the actual percentages are subject to change and only represent a best guess from 23andme. I do know what generation each of my major contributors came from - and the results are...close.

In general, indigenous people and their markers are among the most difficult to put into the algorithm. Since the two daughters have a mostly indigenous mother - do they at least have the same mtDNA?
 
  • #819
"If a suspect is arrested and charged with a criminal offense after an FGG profile has been entered into one or more DTC services, the investigative agency shall make a prompt formal request that all FGG profiles and associated account information and data held by any such service be removed from its records and provided directly to the investigative agency.
[BBM]

This is exactly what they did. Follow protocol. BK was arrested and charged and the profile was removed as required.

P. 6

"[In line w DOJ policy] [O]nce Defendant was in custody, the FBI removed the SNP profile … and no longer has access to view much of the information it used to create the family tree and cannot view it again without resubmitting the SNP profile to the genetic genealogy service(s)."


The investigative agency shall document its request and compliance by the DTC service(s). All FGG profiles, account information, and data shall be retained by the investigative agency for potential use during prosecution and subsequent judicial proceedings."

It is stated they did

Page 6

"To the State’s knowledge, the only records that reflect the FBl’s efforts to create Defendant’s family tree is the family tree itself, notes jotted down by FBI agents as they constructed the family tree, and any records created to document the removal of the SNP profile from the genetic genealogy service(s) pursuant to the DOJ Policy. The State has not seen—nor does the State possess—these records or copies of these records."

Insofar as account information, it has been stated they did not download it so therefore, they had nothing to retain except:

1. The Defendant's family tree
2. Notes jotted down by FBI agents as they constructed the family tree
3. Records created to document the removal of the SNP profile from the genetic genealogy service(s) pursuant to the DOJ Policy.


Page 4

"The FBI built the family tree using the same tools and methods used by members of the public who wish to learn more about their ancestors. For example, the FBI consulted social media, viewed vital records such as birth and death certificates, and viewed other information already contained in the user portal for the genetic genealogy service, including unverified information submitted by other users of the genetic genealogy service."


I read this as (MOO) "The defense is asking for things that do not exist. In order to "make" them exist you'd have to order the FBI to do the work all over again. Issue warrants (if they were required), input the profile again, and do the work all over again - just to "take notes" for the defense. (Page 6 " ... the FBI ... cannot view it again without resubmitting the SNP profile to the genetic genealogy service(s).")

We don't think they're entitled to it under the rules but if you find they are let's have a private meeting where you could look at the information and see if it's sufficient to meet the demand, and if it must be handed over, that it be handed over under a protective order." ["In the alternative, if the defense can establish that IGG information is relevant, the State asked the Court to conduct an in camera hearing so the State can present information related to the IGG information and enter a protective order pursuant to I.C.R. 16(1)." https://s3.us-west-2.amazonaws.com/...fendants+Third+Motion+to+Compel+Discovery.pdf]




jmo
 
Last edited:
  • #820
My own opinion is that there is no way of "documenting" the mental processes behind IGG. It's simply impossible.

On the one hand, it's technically easy. Even if there had not been somehow named Kohberger in Othram's database (or whatever database it was), there would still be matches, because we are all related. Naturally, the higher the percentage of match (let's say 25%), then it's a certainty that the two people in question are related. That's because, aside from point mutations, everyone's DNA is sourced from just one place (their parents, then their grandparents, then great grandparents and so on). The mathematics (and SNP choice and chromosomal span) of this process is not difficult, easily accomplished by a computer. Is the computer "thinking"? What is there to report about what the computer does?

Many of our own human cognitive functions are easily replicated by machine; some are not. Obviously, a human could sit down with 600,000 data points on a long read out and try to personally verify that two different samples match and at what percentage/degree of confidence. That human had better find the same thing the computer does, though, because when it comes to comparing long (apparently random) lists of letters, computers are awesome.

Those of you who have taken biology know that one-half of these letters must mirror the other half - so the 1.2 million nucleotides in those 600,000 SNP's can in fact be halved. Super. Still quite a chore - and the repetitive nature of the task - make it more prone to human error.

Here's the type of chart that emerged from Idaho State Labs after putting the sample from the sheath into a DNA sequencer (for one in action, go to youtube and find Spencer Wells's The Journey of Man and skip to about 45 minutes, in the part where Wells is seeking one rare and specific SNP in India (an SNP that heretofore had only shown up in the Australian aborigine population - and it's my understanding that all of the Australians had that rare SNP).

View attachment 444356

This data is part of someone's genome. Make it way longer and that's what was sent to Othram, IMO. That's how DNA data is recorded, everywhere on Earth that I know of (because each letter stands for a chemical term - like guanine - and all biochemists use the same terms.

A machine then looked through their database (like the search function here on Websleuths) and found a significant match. Match means "is identical to" a certain segment of this data. Since Othram has chosen the 600,000 SNP's most associated with individual variation in *advertiser censored* sapiens, and since geneticists know the relative distance of two samples by statistical methods, this is quite effective at finding our relatives. I have 5000 relatives on 23andme. Some of them are descendants from 6xGreat grandparents, of course. But the system is sensitive enough to detect that. One doesn't even need to write anything down - the computer does the "writing," and using pen and pencil would be silly and impractical. The computer does the "thinking" and "recording."

Kind of like the chambers on a slot machine clicking into place - with just 4 symbols (ACTG) repeated 600,000 times in a specific sequence (your own DNA is as specific as it gets - unless you are an identical twin).

If the Defense was asking the Kohberger family to reveal whether Bryan has an heretofore unknown identical twin, and asking them to produce evidence that such a twin did NOT exist - it would be similar. One can MATCH things (children can do it), but to say what is NOT present in a pattern is quite difficult - esp. if the thing NOT there doesn't exist in the first place. In the case of DNA, it would be obvious if something other than ATCG were there (but I've never ever seen a computer make a mistake and put a wrong letter in - as they are only given FOUR to work with).

IMO.
What the defense is talking about is their hope the police used a profile from someone that opted out to build the DNA family tree from the DNA profile on the sheath.
 
Status
Not open for further replies.

Members online

Online statistics

Members online
114
Guests online
1,642
Total visitors
1,756

Forum statistics

Threads
632,351
Messages
18,625,142
Members
243,101
Latest member
ins71
Back
Top